#i just needed 1k more tokens

[phantaloon]

yall will not get it

but im so close and so fucking far to getting khasos

I NEED HIM TO ASCEND MY ANTANDRA

I'VE BEEN STUCK WITH MOST HEROES AT 180 BC IM NEVER ABLE TO ASCEND MY ANTANDRA

FUCK

[thewickedjazzy]

Could u do dazai x reader with ocd?

Featuring: Osamu Dazai x gn! ocd reader

Content: sfw, fluff, slightly angst, mention of some swearing words, I believe that's it?,

Author's note: hiiii anon, so sorry for the late response, I swear I was working on it since I received it. I'm also working on all of yall's requests - I'm not sure if you wanted a fic, drabble, or headcanon so I just did a rock paper scissors.

Word count: 1k

❥ Dazai is a smartass, quick witted even, and don't forget that he works at the agency as a detective so of course he will quickly notice how you constantly worry about things going wrong—whether it’s leaving your water bottle uncapped or making a mistake at work. He sees you zoning out at the office, caught in a loop of anxious thoughts, and quickly deduces that it’s more than just everyday stress.

❥ At first, before really getting to know you, he was just a quiet observer. But he never judged, not even once. He’s got his own mental shit to deal with, so he knows how fucked up minds can be. He’d make mental notes of your quirks, just in case he needed to handle them during a mission or whatever.

❥ As he gets to know you better, he figures out how sweet and lovely you are—unlike the twisted side of your mind. It’s like an addiction for him; he craves that balance. He enjoys being friends with someone who has their own dark side, just like he does. Maybe you’ll understand where he's coming from, push him to be a better version of himself.

❥ It’s only after your relationship becomes official that he feels entitled to meddle in your problems. He believes that getting involved in such matters is only appropriate once there’s a solid, established connection between you. Until then, he's just an observer.

❥ You were the one who approached him first, explaining that you have OCD. You felt more anxious than you’d anticipated, having considered telling him many times before but never managing to do so because your mind held you back. He remained calm, and it was only when he revealed that he had known all along that you felt embarrassed. In his typically laid-back manner, Dazai might say, “Oh, I’ve known about your OCD for a while now. You don’t need to be embarrassed—it's actually quite endearing. It doesn’t change a thing between us. We’ll work through it together, and honestly, it’s no big deal.”

❥ One day, while on a mission with him and Atsushi, you were deeply focused on analyzing the crime scene. You managed to uncover evidence that even Dazai had missed, piecing together how the victim was killed with remarkable accuracy.

❥ Needless to say that dazai was surprised by how quickly you cracked the complex case. Even after the murderer was arrested, you continued to ramble about additional details, anxiously checking the file in your hand and worrying about missing something, though there was nothing left to find. Eventually he stopped you, gently holding both your wrists. As you looked up, apologising profusely, he simply shushed you, saying, "It's fine. The case is solved, and we’ve caught the murderer."

❥ Atsushi the poor kid he is watched the exchange between you and Dazai with curiosity. When you reluctantly handed him the file with a nod, Atsushi leaned in and whispered to Dazai, asking if you were alright. Dazai simply smiled and replied, "They're just the perfect detective, doing their job flawlessly."

❥ Dazai is surprisingly attentive to your needs when it comes to your OCD. If you’re stuck in a loop of checking something multiple times, he’ll gently steer you away from it by engaging you in conversation or distracting you with a task. He might even playfully mock himself to make you laugh, using humor to ease your anxiety. “Looks like I’ve accidentally left my keys at home for the fifth time this week. Isn’t that a catastrophe?”

❥ If you’re particularly anxious about a big case or an important event, Dazai might surprise you with a comforting gesture. He could leave little notes or small tokens of reassurance in places you’d find them, like a hidden message in your desk drawer or a comforting snack in your bag.

❥ When you’re feeling overwhelmed, Dazai has a knack for knowing when you need space and when you need company. He might give you space to deal with your thoughts but always makes sure you’re not alone. He could show up with your favorite coffee or a calming playlist, knowing just when to intervene.

❥ Dazai might not always express it directly, but he has a deep understanding of how your OCD affects you and often subtly adapts his behavior to support you. He might avoid making sudden movements or loud noises if he knows it could trigger your anxiety. His sensitivity to your needs shows how much he cares.

❥ Dazai would likely offer reassurances in his own, less conventional way. Dazai might not be that sentimental boyfriend but he makes casual but heartfelt statements that affirm his love, such as “You’re the only one who can put up with my nonsense,” or “I can truly be myself with you, without a second thought.”

❥ Dazai would be more mindful of his humor, staying away from jokes or comments that might trigger your OCD. He might use humor in a way that’s supportive and aligns with your sense of comfort rather than distress.

❥ If someone makes fun of you, Dazai would likely handle it with a blend of nonchalance and protective instinct. He might casually brush off the mockery with a witty comment, like, “Oh, they’re just jealous of your intellect,” showing his confidence in you. If needed, he would also step in to defend you.

➵Want more of Osamu Dazai ?

[httpsghostie]

Hi!!! I'm obsessed with your writing, especially all the Ghost stuff. YUM!! I was wondering if you'd be able to write an angsty and emotional blurb with DBF Ghost...Maybe they get into a fight or something

Favorite Regret

pt 1 / pt 2 / pt 3

took me a while to get to this but hope you enjoy it <3

any similarities with sleep token is a mere coincidence I am not responsible for this

Summary: Simon doesn't know how to handle love, so he pushes you away.

Word Count: 1k

Warnings: dbf!Simon 'Ghost' Riley x f!reader, angst/hurt, alcohol abuse, age gap (reader's in 20s and simon in his late 30s) // masterlist

Your story with Simon began a few months back when he returned from his duty and desperately – more like unconsciously, needed affection.

It happened, you were there, emotionally available for you dad’s best friend, it was wrong, he knew it, but Christ you were too good to be true.

Always there for him, dropping off work and studies only to take care of a fever he’d have, pampering him with lots of kisses, making him a warm cup of tea. He was never a physical touch kinda guy, but with you, the way you felt lying peacefully on his chest, he couldn’t help it, he became clingy, almost annoyingly clingy.

And he filled you with empty promises, painting a future for the two of you. He held you in his arms, he let you invade his personal space – which he then called our personal space. You were in fields of elation, but every color seemed to fade as he shut the doors to his heart again.

It’s almost like he determined you were no longer useful, you had given everything to him, you trusted him wholeheartedly, just for him to take it all from you and throw it in the bin.

You hated him, the image of the one you loved and cared for turning black and white as the entirety of your relationship turned meaningless. It’s like he gave you a whole beautiful horizon that stretched ahead in front of you, and then made it disappear from your view, leaving you standing at the edge of uncertainty.

He took what he wanted from you and left.

“You’re not getting it, this is never going to work, your dad doesn’t even know about us, he’ll kill me if he finds out. God, what have I done?” He had his back turned to you as he sat on the edge of the messy bed from your previous actions. You covered your bare chest with the blanket, your face getting red and hot.

“No, you don’t get it, Simon. I’m a fucking adult, and you are a childish, self centered idiot, can’t you see it? You just cared for me when you needed someone to be by your side, and now that you don’t need me anymore, you’re gonna throw what we had away because what? You said so?” You screamed back at him, eyes filled with hot tears that dared to roll down your cheeks.

“Call me what you want, I don’t care. This was a mistake.”

“A mistake? Is that what you’re gonna call me?” Your voice was shaky, and you tried your best to keep your cool. You dressed yourself quickly, still expecting an answer from him, but he just sat there, watching all the events unfold. “What is wrong with you?”

“You don’t wanna be with me, I can’t even stand myself, you deserve someone better.”

“Someone better?” You stood in front of him in denial, crossing your arms, but he never looked at you. “Did I just hear you right?” 

“I’ve made myself clear, kid. You should go.” He finally got up, and your eyes accompanied his emerging size. But he still never looked at you, he just walked up to the front door to his apartment and opened it. You slammed it closed, pointing a finger to his face.

“You listen to me-”

“No, you listen to me.” He grabbed the sides of your arms, holding you firmly. His teary eyes staring down into your soul. “Do you think this isn’t hard for me, either? You don’t understand, we can’t be together, I can’t disappoint your father like that, I can’t be with you. I should’ve never texted you in the first place.”

And the way he spoke made you inhale a familiar smell escaping from his lips. Whiskey.

“Simon, are you drunk?” You cried. 

“Go.” He finally let go of your arms, leaving your skin almost bruised.

You brushed your arms and headed to his kitchen, trying to find the bottle that was shoved in the trash can. You picked the empty whiskey bottle, noticing a discarded pack of cigarettes — empty, too, and smashed.

"You drank half a bottle?" You raised the object to your head, he looked down, rubbing his eyes. "Unbelievable."

"You're still young, you have a prosperous future ahead of you and I won't be the one to hold you back." He said quietly, fighting back the tears and the agonizing pain in his chest.

"God, you talk like you have a terminal disease." You scolded, leaving the bottle on the kitchen aisle and grabbing your stuff from the couch. "Maybe you do, Simon, maybe your terminal disease is not being able to accept love. Maybe your terminal disease is you becoming the one who you swore you'd never become."

It came out strained, and you didn't realize you were comparing him to his father until his hand gripped your wrist tightly. He looked deep into your eyes, and you could see how pissed he was.

"I'm not him. Get out." You pulled away and he opened the door one more time. 

You didn't comply, your eyes desperately tried to find the loving Simon you knew, the one who held you close in the night, the one who couldn't stand your cheesy movies but still watched them with you. But he wasn't there, the version you saw in front of you had killed the previous one. The Simon you knew was dead.

"Simon, I didn't-"

"Get out." 

Now he screamed, almost spitting out, his hot tears flooding his eyes and rolling down his red cheeks. 

He stood outside the apartment watching you clumsily press the buttons on the elevator, giving him one last glance before finally walking away.

A mistake. That's how he said it. This relationship was a mistake, and you were his favorite regret.

taglist: @butterbunana @snoisisabitch @nuhteyam @iamabsolutelynothere @blissful--moon @jellyluvr @khomugi @xaintxun @kichimiz @frog-spot @sasukeswife3 @aly0be

[jcxbliss]

“Tough day”

Pairing - Yoon Jeonghan x Female Reader 

Genre - Fluff, Idol AU, SFW, established relationship au, idol x normal person

Word Count - 1k

Author's Note - Cute little piece<3

(ANYONE CAN JOIN THE TAGLIST)

Tags -@hipsdofangirl

The soft light of the late afternoon sun filtered through the large windows of the cozy living room, casting a warm, golden glow over the space. You perched on the edge of the sofa, a gentle smile playing on your lips as you glanced at the clock on the wall. It was a few minutes past the hour when you had expected Jeonghan to arrive, and though you knew he was likely running late, your heart fluttered with anticipation.

Today had been particularly grueling for Jeonghan. His dance practice had been intense, and his usual charismatic energy had been dimmed by the exhaustion of the day. You could picture him backstage, his face flushed and tired, but still carrying that undeniable charm that had won your heart. You had always admired his dedication to his craft, and despite the challenges he faced, his passion never wavered.

With a sigh of contentment, you reached for the small, neatly wrapped gift you had set aside on the coffee table. It was a thoughtful token of appreciation—a cozy, oversized hoodie that you had picked out for him, one that you hoped would provide comfort and warmth after his long practice. You imagined how he might bury himself in its softness, finding solace in its embrace.

As if on cue, the sound of footsteps echoed from the hallway. Your heart skipped a beat as you heard the familiar jingle of Jeonghan's keys followed by the soft creak of the front door opening. You stood up, smoothing your dress and taking a deep breath, your excitement bubbling just beneath the surface.

Jeonghan entered, his usual bright smile replaced by a tired but genuine grin. His hair was slightly disheveled, and he wore a light sheen of sweat on his brow. As soon as he saw you, his face lit up with relief and affection. He immediately closed the distance between you, wrapping you in a gentle, yet tight hug.

“Hey,” he murmured, his voice slightly muffled as he buried his face into the crook of your neck. “I missed you.”

Your heart swelled at his touch. You ran your fingers through his hair soothingly, feeling the tension slowly melt away. “I missed you too,” you replied softly. “I have something for you.”

Jeonghan looked up, his eyes glinting with curiosity. You handed him the gift, and he took it with a grateful smile. His eyes widened as he unwrapped the hoodie, and a soft chuckle escaped his lips when he saw the oversized garment.

“This is perfect,” he said, slipping his arms into the hoodie and hugging it around himself. The softness of the fabric seemed to lift a weight from his shoulders, and he closed his eyes for a moment, savoring the comfort. “Thank you. I needed this more than you know.”

You watched him with a tender smile, your heart warmed by the sight. You could tell that the hoodie was already becoming a source of comfort for him. “I’m glad you like it,” you said. “I thought it might help after today’s practice.”

Jeonghan leaned in and kissed your forehead, a gesture of both gratitude and affection. “You always know just what I need,” he said, his voice filled with sincerity. “I don’t know what I’d do without you.”

You both settled onto the sofa, Jeonghan pulling you close as he relaxed into the cushions. With the hoodie enveloping him and you nestled against his side, he let out a contented sigh. The exhaustion from the day seemed to dissipate, replaced by the warmth of your presence and the comfort of the soft hoodie.

The golden light of the setting sun cast long, warm shadows across the room as you and Jeonghan settled into the sofa. The soft hum of the city outside was a gentle backdrop to the comforting cocoon you had created together. With his head resting against your shoulder and the oversized hoodie enveloping him, Jeonghan seemed to finally let go of the day’s stress.

You gently stroked his hair, savoring the peaceful moment. “So, how was practice?” you asked, your voice soft and curious.

Jeonghan sighed deeply, a mixture of relief and exhaustion in the sound. “It was really tough today,” he began, his voice muffled slightly against your shoulder. “We were working on a new routine, and everything had to be perfect. I felt like I was stumbling through the whole thing.”

You could hear the strain in his voice, and you tightened your hold on him comfortingly. “I’m sure you did great. You always do. But I understand it can be draining.”

He chuckled softly, though it was tinged with fatigue. “I appreciate that, but today was just… different. There were so many elements to juggle, and by the end, I just felt so worn out. I kept thinking about how nice it would be to come home and just curl up with you.”

Your heart melted at his words. “I’m glad you’re here now. You deserve to relax after such a demanding day.”

Jeonghan shifted slightly to look up at you, his eyes soft and sincere. “You know, the whole time I was practicing, I kept imagining coming home and cuddling with you. It was the only thing that kept me going. I knew that, no matter how tough things got, I’d have this moment to look forward to.”

You smiled, gently brushing a lock of hair away from his face. “You’ve got me here, and I’m all yours. We can stay like this as long as you need.”

He sighed contentedly, closing his eyes as he settled back against you. “I really don’t know what I’d do without these moments. They’re my sanctuary.”

You could feel the tension in his body melting away as you continued to stroke his hair. “And you’ve got plenty more moments like this to look forward to. We can just enjoy the evening, talk, or even just be quiet together.”

Jeonghan’s lips curved into a small, relaxed smile. “That sounds perfect. Sometimes, I think this is exactly what I need the most—a simple evening with you, where nothing else matters but being close.”

As the room grew darker, the soft glow of a nearby lamp illuminated the peaceful scene. You both let the silence wrap around you, a comfortable and soothing presence. Occasionally, Jeonghan would murmur something about his day, but mostly, he seemed content to simply be with you, wrapped in the warmth and security of your embrace.

You leaned your head against his, feeling his steady breathing and the gentle rise and fall of his chest. The world outside seemed to fade away, leaving just the serene bubble of your shared space. The day’s troubles and exhaustion seemed distant, replaced by the calming rhythm of your togetherness.

In this quiet, golden evening, surrounded by the soft embrace of love and comfort, you both found a peaceful haven—a perfect end to a challenging day, where all that mattered was the closeness and warmth you shared.

[spoiled-milk]

dan heng as your roommate (modern au)

a/n: in compliance with dan heng’s five star release being yesterday here’s brainrot for my beloved also praying that those who pull for dan heng win their 50/50 or get him early (it took me 80 pity with guaranteed 😭😭). will be mia so i can farm more stellar jades for his weapon since he wanted to come home on hard pity

content warning: mentions of alcohol consumption, other than that none that i can think of. no mention of reader's gender

word count: ~1k words

you share a small apartment with dan heng. you used to share it with one of your friends, but a sudden job opportunity for them turns into you frantically finding a roommate to fill so that you’re not paying the full price of the rent next month

you find dan heng through one of your friend’s friend. March (your friend) knew someone with grey hair who then gave dan heng your details since he was looking for a place to live

you two officially meet at a coffee shop where you discuss rent, house rules, and the lease and you did not expect him to be hella attractive

rules are as follows: there’s a chore list on the fridge in the form of dry erase board, if bringing over friends you must let the other know before you bring them over, no going into the other person’s room without their knowledge, and keep hooking up to a minimum if possible or keep it quiet

at first he seemed like a quiet guy who wanted nothing to do with you which like didn’t hurt your feelings too much yk you just needed someone to pay half of the rent, but slowly he starts to open up to you and you have no problems opening up

usually you take turns cooking, but he’s come to realize you only know how to make is boxed food and breakfast foods, so after eating waffles for the 12th time that week he decides he’ll cook dinner for the most part

dan heng without fail will always give you the last piece of anything he makes. dumplings? you can have the last one. 12 pack of juice in the fridge? he won’t even go get more unless you drink the last one. you make brownies as a token of appreciation for him and insist you don’t want any and all of it is for him? you find the last piece tucked away in the fridge with your name on it

he makes it a point to do the dishes together. at first he argued that he could do it but after much begging from you, he allows you to help him rinse the dishes

this has become a nearly every night thing. you eat whatever dan heng makes then y’all do the dishes. most of the time you’re doing the talking, but it’s not exhausting to talk to him like you talk to others

you don’t know much about his past but he sure does knows a lot about your past. you’re sure march probably unintentionally spilled about your past, but you don’t really mind knowing it’s just dan heng. march has told you to ask him but when you do, he softly shuts it down

it doesn’t hurt your feelings of course. you understand that everyone has their secrets and no one is entitled to his, but it does hurt a little that he doesn’t seem like he wants to tell you since you thought you were getting pretty close as friends

oftentimes at the end of a stressful week, you find some movie on some streaming service and crash on the couch with snacks littering the coffee table. sometimes dan heng will come join you on the couch to whatever movie you put on, regardless of genre. rom com? he’s got his eyes glued to the screen. some environmental documentary about the gas leaks that effect a underprivileged community? he’s sat next to you nodding his head to the tv like a middle aged dad

most of the time you fall asleep in the middle of these movies and most of the time when you wake up in the middle of the night you either find yourself covered with a fluffy blanket or you find yourself in your bed with no memory of how you got there. you suspect dan heng carried you there, but you’ve never had the courage to ask him

there was one time when you woke up from sleeping in the middle of the movie and the tv was left on to the end credit scene. you reach for the remote next to you and turn off the tv. you start to shift a little but then you notice a weight next to you, it’s your roommate. he’s got an arm around your waist and he’s pulling you closer to him

you’re too tired to properly comprehend the situation so you let it happen. you lean your head towards his chest and end up falling asleep listening to dan heng’s heartbeat. when dan heng woke up that morning realizing what had happened, he could not look you in the eye for a week without blushing

he also takes care of you whenever you come home shitfaced from an outing or get shitfaced by yourself with 3 bottles of soju. it doesn’t happen often and you try not to drink often, but when you decide to go out, dan heng somehow knows that you’ve been out drinking and will always be at home waiting for your return

your friends drop you off at your apartment and dan heng helps you into bed. he takes a baby wipe and wipes off any dirt on your face. he’s so gentle with it that it honestly feels like a dream. you lowkey cherish it every single time, it feels so intimate that you wish he meant it in that way and not as just a friend

just as you’re losing consciousness, dan heng gets close to you but your vision is slightly blurring so you’re not sure what he’s up to, but you trust dan heng 100% to not take advantage of you

as you scrunch your eyebrows you feel something soft press against your cheek and then once again on your forehead. they’re brief and short, but you can feel your body get hotter and hotter, not even sure if it’s the alcohol or dan heng (or both maybe)

you pray that your big crush on him isn’t obvious but march tells you otherwise. so now you only hope that dan heng can’t tell but honestly it’s really hard to read his feelings, but you’re not sure how long you can keep up this “he’s just my roommate” act to your friends and yourself

[teacasket]

kiss kiss, fall in love

genre: fluff au: academy au, light academia au warnings: swearing word count: 1k   pairing: gn!reader x kim seungmin

Seungmin, to put it mildly, is fucking rich. To be fair, about 99% of the students attending the prestigious Clé Academy are, but as the heir to a pharmaceutical conglomerate, he’s the top one percent of the one percent. Every time you buy cold medicine or even socks from the market in the cheaper part of town, you’re reminded exactly how much money he has in comparison to you.

You’re here on scholarship; nothing more needs to be said. 

Despite the differences—and background resentment of how fucking wealthy and entitled the student body is at times—you find yourself mostly okay with Seungmin. Due to seating arrangements in biology class, he winds up being your lab partner and by proxy, your study partner as well. Kind and intelligent, he’s better than a lot of your previous group members. He pulls his own weight and is punctual with the lab reports, so he gets your stamp of approval, though his privilege does show with the occasional out of touch with reality comments. No, not everybody has a family lawyer on retainer. But again, to be fair, 99% of the students do, so you’re sort of the one who seems out of touch with reality. Their reality.

On the morning of November 11, you are once again confronted with that reality. He interrupts your early morning review session inside the library with a white pastry box from a bakery that you know has an extremely long waitlist for desserts.

“Happy Pepero Day,” he says as he sits in the chair across from you. “This is for you.”

“Happy Pepero Day,” you reply back. You didn’t foresee this happening, so you don’t have any available boxes to give him. “Thanks, but I don’t have anything for you.”

“That’s okay. Open it.”

Your stack of notebooks and flashcards form a dividing line along the table, so you have to awkwardly stand up and reach for it. Under his anticipatory gaze, you pop open the lid of the box to reveal some very adorable macarons. At first, you think that the baker made a mistake, but you quickly realize that the four interconnected circular shells imitate Pepero sticks. They’re all in various pastel shades and have animal faces, and you have never wanted more to melt into a puddle. Telling Seungmin about your love for all things cute, was a mistake.

“Do you like it?” he asks when you can’t do anything else but stare at how darling the macarons are. “I didn’t know what you would want, but I guessed vanilla, strawberry, and lavender from your favorite ice cream flavors.”

Fuck Clé Academy for having soft serve machines available at lunch.

“Thanks, but I…”

Can’t accept this because I will feel indebted to you even though this is just a token of appreciation.

“I’m allergic to gluten,” you decide.

A second passes before Seungmin laughs and says, “No, you’re not. I see you stealing non-gluten-free cookies and sandwiches from the cafeteria all the time. Besides, even if you were, macarons don’t have gluten. Try again.”

Fuck Seungmin for being so attentive.

“They’re expensive, okay? It feels weird to me to accept it, especially since we’re just lab partners.”

“Just lab partners?”

“Friends,” you amend. “Look, the point is, this is really nice of you, but this box probably costs the same amount as our tuition, and I’m not comfortable with that.”

“What if I was your boyfriend? Would that make you feel better?”

You can only blink at the audacity. If you had this much confidence, you would have taken over the world by now. “Well, you’re not. And no, it wouldn’t. Seriously, what are you trying to get at? I banned Chaeryeong, one of my close friends, from giving stuff like this to me. You’re not exempt from this rule.”

For some strange reason, Seungmin laughs. You’re fully flushed from embarrassment or mortification or plain anger because why is this hilarious?

“Did you read the note?” he asks, stifling a grin.

“What note?” You check the box again, but it holds only pastries. “Was there supposed to be one?”

His face drops alarmingly quickly, so you spin the box around him so he can see for himself. This time, it’s Seungmin’s turn to be flushed. It’s not a delicate rosy pink blush that you expect all “well-bred” individuals to have but an uncontrollable wildfire. It’s almost endearing to see it on Seungmin, the most composed person you know.

“Close your eyes,” he says as he holds up a blank pink flashcard. “Also, can I borrow this?”

You resist the urge to tease him in this crisis. You told me to close my eyes. I don’t know what ‘this’ is. “Sure.”

You hear the click of a pen and then some scratches on paper. After a minute, Seungmin tells you to open your eyes.

The pastry box has changed—there is now a pink flashcard with the words “Can I be your boyfriend?” on top.

“I just told you that it wouldn’t change my mind,” you say while you figure out an appropriate response. Fuck. You don’t like him that way, but you can’t risk hurting his feelings. You’re lab partners. What if he suddenly decides he’s so offended that he stops doing his share of the lab reports?

“Is this a rejection?”

“Try again next year,” you distantly say as you gather your things, being sure to leave Seungmin’s gift behind. “See you in bio.”

Both of you are great at pretending everything is fine because you manage to remain friendly with him and he treats you in the same manner as before. The following year, you have honors chemistry with him and become lab partners again. To your terror, everything becomes not fine when you catch yourself staring at his profile during a midterm review session. You can rationalize it all you want, but you can’t ignore the sudden onslaught of butterflies you get when he listens intently to your explanation of magnets.

Fortunately, on November 11, Seungmin does, in fact, try again, this time with a simple pack of strawberry Pepero and a lot less presumption. And this time, you accept him and the excited kiss he places on your lips.

[thatsthewrongwallcraig]

OKAY HEAR ME OUT!! cunnilingus with overstimulated reader and subby kappa who has an oral fixation and is basically super pussy drunk and eating the reader out is pretty much his favorite activity (like man could and would go for hours)

You know where it's at, nonnie! 🥴🥴

Worship II

Summary: Kappa thriving on his oral fixation.

Pairing: Kappa x fem!Reader

Word Count: -1k

Content Warnings: Smut 18+!, Kappa Being Such A Massive Service Sub, Kappa Being A Very Good Puppy, Oral (F Receiving), Tongue Fucking, Praise Kink, Affectionate Degradation, Talk About Collaring, Aftercare

A/N: Sometimes babygirl is a deranged cult leader 🤷🏻‍♀️

Tagging the horny horde:

@crypticsewerslut @quicksilversg1rl @alalalaaallaaalaaa @star-milk-tea @angelsanarchy @milsthouqhts @amayalul @roryculkinsbf @roryculkinsgf @spookyorchid @arch1viste

You've got my body, flesh and bone

The sky above, the Earth below

Nothing to say and nowhere to go

A taste of the divine

- The Summoning By Sleep Token

With your ass propped up on his thighs, your shaky knees dangling over his shoulders and his face buried deep in your almost aching cunt, you bit down hard on your lower lip.

"Fuck, Kappa, I'm going to turn sore when you keep sucking my clit like that." You groaned under your breath, the overstimulation easing off a little as drew his head back for a moment, his gaze searching yours.

"I know you just want to be a good puppy for me, don't you? But you gotta be a bit more careful, okay?" His plush, pink lips were covered in a shiny layer of your past two orgasms down to his chin as he looked at you with lust-blown pupils.

" 'M sorry, Ma'am. I got a bit carried away with it, no?" For a brief moment he gnawed at the inside of his cheek with his brows arched to a soft frown.

"Yeah, yeah you did, but don't you worry about it. Just be gentle." You chuckled softly, giving him a short nod to indicate that you wanted him to continue.

"Will do, Ma'am." He answered with a faint smile pulling at the corners of his lips before leaning back in again, the tip of his tongue now caressing your fairly overstimulated clit with careful kitten-licks, nudging it softly in between gentle kisses.

"That's it, Kappa." It spilled out of your mouth in a breathy moan, your sensitive nerves tingling with every touch of his tongue.

"Fuck, you really love eating me out, don't you? Such a perfect little puppy boy for me. With nothing else on his mind than to be of service." You looked down at the frizzy mess of his black hair between your thighs, followed how his face moved against your cunt and how his breath hitting your slik-coated folds sent goosebumps across your body.

"Uh-Huh…" His response got lost in your skin as he pressed another gentle kiss to your clit, his lips gingerly enclosing it.

"How long have you been worshiping my pussy now? 2 hours, maybe even longer?", You choked back a guttural groan as his lips hit that one overly sensitive spot in particular, "You've got such a pathetic, needy oral fixation. Always kissing, licking or suckling at me somehow like the perfect fucking attention whore that you are."

You felt him mewling into your cunt as the words pinched him right where they should.

"Craving my attention all around the clock, no? And getting oh so excited when I decide that you have earned yourself some.." You pushed further, trying to talk yourself out of the bubbling and boiling mix of overstimulation and your body getting ready to come undone for a third time.

The fine line of burning arousal and equally prickling, biting overstimulation fading further with every touch of his and you couldn't decide if you needed him to stop right nor or let him push you past the threshold. Instead, you instinctively squirmed and wiggled your ass, not exactly sure whether this was causing more friction or easing it up.

"Good God, Kappa, fuck!" I shot right out of your mouth as you involuntarily arched your back, your thighs caging his face between your legs whilst your fingernails drilled themselves into the sheets.

For a split second you were actually worried about choking him out for good but that thought, alongside anything else in your mind, got sucked into oblivion as the sore coil in your lower abdomen snapped, spilling your release all across his face as a croaked, almost pained moan erupted from the depths of your lungs and yet Kappa took it up a notch by shoving his tongue into your cunt right as your body started trembling. Your walls throbbed and contracted around his tongue as he tongue-fucked you through your orgasm and you forgot how to breath for what seemed to be an eternity.

"Fuck, Kappa." You whined, only slowly coming down from your orgasmic high as he gently slid your knees from his shoulders and pulled himself out of you with a wet squelch, his entire face lighting up in a blissed out expression as he looked at you with nothing but admiration in his eyes, your chest heaving up and down.

"Come here." You opened up your arms, inviting him into a tight embrace.

Following it, Kappa crawled over to you, resting his head on your chest and sighing softly as you pressed an avalanche of kisses to his forehead.

"You can't imagine how much I fucking love my perfect, little puppy boy." ,You raised your hand to the back of head to pet him carefully, "Getting me off again and again when I want you to.."

He nuzzled his face into the crook of your neck with a satisfied but thoroughly tired smile.

"You know, the other day, I thought about getting you a pretty, little collar. Maybe with a tag on it too so that everyone can see just how much I love you. Would you like that?" You noticed Kappa shifting a little, his body pressed close to yours.

"You'd do that? Really?" He whispered into your neck, his voice laced with excitement.

"Oh, absolutely. My perfect puppy deserves a pretty collar, doesn't he?"

[mistydeyes]

OMG THERES 1K OF YOU!! I cannot believe it! Thank you all so much for all the love and support these last few months :) Literally tearing up at how many people love my works. It all started with a silly little pharmacist fic and look at where we are now!

To celebrate, I will be opening up my inbox for some prompts! For the next week, feel free to send me a character, a prompt below, and any other details (ex. female, male, or g/n character) :)

aesthetic prompts 💌 - CLOSED NOW! thank you for all the submissions :)

soft and cute - all fluff!

vase of flowers - they return home to your door with a small bouquet of flowers with rudy

loveseat - shopping for furniture for your shared home

linen tablecloths - hosting a dinner party (one that is pinterest worthy)

orange juice - sick days call for the best care

vintage maps - a spontaneous trip with an event filled itinerary

cherry lip balm - what was your first kiss like? with graves

paper rings - “put this on” “why did you make a ring out of the straw wrapper?”

ramekins of soup - you both are not experienced chefs but you try a new recipe in the kitchen

tuberose- looking at old photo albums and reminiscing on old memories

tote bags - “hey can you hold this for me?” “why do you have so many things?” “it’s not my fault these pockets don’t hold anything”

sugar cubes - they surprise you with your favorite dessert/sweet treat with price

music - “i made you a playlist!”

pearls - a fancy dinner out with expensive clothes and dining

folded sheets - "hi, i'm your neighbor across the hall :)"

gold - a small gifted ring as a token of your love with price

unlit cigarette - meeting in the most unexpected of places

hot chocolate - a warm snowed in cabin with nowhere to go

house plant - “for your new apartment!” “you know i’m not here enough to water that”

pastel bonnets and rattles - “we are NOT naming our child that”

perfume - perfume bottles decorate your bureau and you decide to let them choose your scent for the day with gaz

friendship bracelets - friends don't look at friends that way, is there something between us? with ghost

shower of tears - angst

crystal glasses - a fight ends with shattered wine glasses and someone is left to pick up the pieces with gaz

running mascara - "i can't believe you would say that to me" with the 141

hyacinth - they decide to interrupt the wedding just as you're about to say your vows with ghost

cardigans - "please don't go, please don't leave me" with ghost

sea salt - you can't do this anymore and just need to get away from them

stars - "do you think in another universe things would have worked out for us?"

cubic zirconia - you both come to the realization that your relationship is beyond repair

cold sheets - "can you hold me just one more time?" with soap and gaz

hydrangeas - a sudden death leaves one of you without the other for the rest of time

cocktails - "you need to stop drinking too much" "what do you know?"

graveyard - "please i don't want to die" with price

enlisted - military au

empty cartridges - "I want you to run, I'll cover you!"

uniforms - they sure clean up well as you look at them in their dress uniform with price

cigarette smoke - you share a cigarette (and exchange indirect kisses) while on patrol with ghost

confidential files - you're the newest addition to the team and you recognize a familiar face

jacket - you forgot to pack you winter gear and they're helping you avoid the cold

angel - this time you're their knight in shining armor as you save them with soap | with price

manila folders - "I'm being transferred to another unit"

bandages - you are the only one with the med kit and are tasked with patching them up

barbells - "you're doing it wrong, let me show you"

MREs - "can we switch? this food is absolute shit"

hit the showers (18+) - a prank ends with you getting your clothes and towel stolen with soap

wigs - you get sent undercover with them

hot and seductive - smut (18+)

lace (18+) - “wait for me in the bedroom, i have a surprise”with soap

stained lipstick (18+) - rough kisses in the back of an uber with gaz | with ghost

stolen glaces (18+) - "are you going to kiss me or not?"

tanning lotion (18+) - you just wanted help with an even lotion application but now their touch is venturing elsewhere

black and white photos (18+) - you take a few photos so they can remember you (and your body) while on deployment with price: part i and part ii

fruit juice (18+) - "let's test out that theory about drinking pineapple juice"

morning sun (18+) - you have nowhere to go and decide to spend it in bed ;) with ghost | with gaz | with price

silk (18+) - “you look absolutely gorgeous tonight”

[snickerdoodlles]

Generative AI for Dummies

(kinda. sorta? we're talking about one type and hand-waving some specifics because this is a tumblr post but shh it's fine.)

So there’s a lot of misinformation going around on what generative AI is doing and how it works. I’d seen some of this in some fandom stuff, semi-jokingly snarked that I was going to make a post on how this stuff actually works, and then some people went “o shit, for real?”

So we’re doing this!

This post is meant to just be a very basic breakdown for anyone who has no background in AI or machine learning. I did my best to simplify things and give good analogies for the stuff that’s a little more complicated, but feel free to let me know if there’s anything that needs further clarification. Also a quick disclaimer: as this was specifically inspired by some misconceptions I’d seen in regards to fandom and fanfic, this post focuses on text-based generative AI.

This post is a little long. Since it sucks to read long stuff on tumblr, I’ve broken this post up into four sections to put in new reblogs under readmores to try to make it a little more manageable. Sections 1-3 are the ‘how it works’ breakdowns (and ~4.5k words total). The final 3 sections are mostly to address some specific misconceptions that I’ve seen going around and are roughly ~1k each.

Section Breakdown: 1. Explaining tokens 2. Large Language Models 3. LLM Interfaces 4. AO3 and Generative AI [here] 5. Fic and ChatGPT [here] 6. Some Closing Notes [here] [post tag]

First, to explain some terms in this:

“Generative AI” is a category of AI that refers to the type of machine learning that can produce strings of text, images, etc. Text-based generative AI is powered by large language models called LLM for short.

(*Generative AI for other media sometimes use a LLM modified for a specific media, some use different model types like diffusion models -- anyways, this is why I emphasized I’m talking about text-based generative AI in this post. Some of this post still applies to those, but I’m not covering what nor their specifics here.)

“Neural networks” (NN) are the artificial ‘brains’ of AI. For a simplified overview of NNs, they hold layers of neurons and each neuron has a numerical value associated with it called a bias. The connection channels between each neuron are called weights. Each neuron takes the sum of the input weights, adds its bias value, and passes this sum through an activation function to produce an output value, which is then passed on to the next layer of neurons as a new input for them, and that process repeats until it reaches the final layer and produces an output response.

“Parameters” is a…broad and slightly vague term. Parameters refer to both the biases and weights of a neural network. But they also encapsulate the relationships between them, not just the literal structure of a NN. I don’t know how to explain this further without explaining more about how NN’s are trained, but that’s not really important for our purposes? All you need to know here is that parameters determine the behavior of a model, and the size of a LLM is described by how many parameters it has.

There’s 3 different types of learning neural networks do: “unsupervised” which is when the NN learns from unlabeled data, “supervised” is when all the data has been labeled and categorized as input-output pairs (ie the data input has a specific output associated with it, and the goal is for the NN to pick up those specific patterns), and “semi-supervised” (or “weak supervision”) combines a small set of labeled data with a large set of unlabeled data.

For this post, an “interaction” with a LLM refers to when a LLM is given an input query/prompt and the LLM returns an output response. A new interaction begins when a LLM is given a new input query.

Tokens

Tokens are the ‘language’ of LLMs. How exactly tokens are created/broken down and classified during the tokenization process doesn’t really matter here. Very broadly, tokens represent words, but note that it’s not a 1-to-1 thing -- tokens can represent anything from a fraction of a word to an entire phrase, it depends on the context of how the token was created. Tokens also represent specific characters, punctuation, etc.

“Token limitation” refers to the maximum number of tokens a LLM can process in one interaction. I’ll explain more on this later, but note that this limitation includes the number of tokens in the input prompt and output response. How many tokens a LLM can process in one interaction depends on the model, but there’s two big things that determine this limit: computation processing requirements (1) and error propagation (2). Both of which sound kinda scary, but it’s pretty simple actually:

(1) This is the amount of tokens a LLM can produce/process versus the amount of computer power it takes to generate/process them. The relationship is a quadratic function and for those of you who don’t like math, think of it this way:

Let’s say it costs a penny to generate the first 500 tokens. But it then costs 2 pennies to generate the next 500 tokens. And 4 pennies to generate the next 500 tokens after that. I’m making up values for this, but you can see how it’s costing more money to create the same amount of successive tokens (or alternatively, that each succeeding penny buys you fewer and fewer tokens). Eventually the amount of money it costs to produce the next token is too costly -- so any interactions that go over the token limitation will result in a non-responsive LLM. The processing power available and its related cost also vary between models and what sort of hardware they have available.

(2) Each generated token also comes with an error value. This is a very small value per individual token, but it accumulates over the course of the response.

What that means is: the first token produced has an associated error value. This error value is factored into the generation of the second token (note that it’s still very small at this time and doesn’t affect the second token much). However, this error value for the first token then also carries over and combines with the second token’s error value, which affects the generation of the third token and again carries over to and merges with the third token’s error value, and so forth. This combined error value eventually grows too high and the LLM can’t accurately produce the next token.

I’m kinda breezing through this explanation because how the math for non-linear error propagation exactly works doesn’t really matter for our purposes. The main takeaway from this is that there is a point at which a LLM’s response gets too long and it begins to break down. (This breakdown can look like the LLM producing something that sounds really weird/odd/stale, or just straight up producing gibberish.)

Large Language Models (LLMs)

LLMs are computerized language models. They generate responses by assessing the given input prompt and then spitting out the first token. Then based on the prompt and that first token, it determines the next token. Based on the prompt and first token, second token, and their combination, it makes the third token. And so forth. They just write an output response one token at a time. Some examples of LLMs include the GPT series from OpenAI, LLaMA from Meta, and PaLM 2 from Google.

So, a few things about LLMs:

These things are really, really, really big. The bigger they are, the more they can do. The GPT series are some of the big boys amongst these (GPT-3 is 175 billion parameters; GPT-4 actually isn’t listed, but it’s at least 500 billion parameters, possibly 1 trillion). LLaMA is 65 billion parameters. There are several smaller ones in the range of like, 15-20 billion parameters and a small handful of even smaller ones (these are usually either older/early stage LLMs or LLMs trained for more personalized/individual project things, LLMs just start getting limited in application at that size). There are more LLMs of varying sizes (you can find the list on Wikipedia), but those give an example of the size distribution when it comes to these things.

However, the number of parameters is not the only thing that distinguishes the quality of a LLM. The size of its training data also matters. GPT-3 was trained on 300 billion tokens. LLaMA was trained on 1.4 trillion tokens. So even though LLaMA has less than half the number of parameters GPT-3 has, it’s still considered to be a superior model compared to GPT-3 due to the size of its training data.

So this brings me to LLM training, which has 4 stages to it. The first stage is pre-training and this is where almost all of the computational work happens (it’s like, 99% percent of the training process). It is the most expensive stage of training, usually a few million dollars, and requires the most power. This is the stage where the LLM is trained on a lot of raw internet data (low quality, large quantity data). This data isn’t sorted or labeled in any way, it’s just tokenized and divided up into batches (called epochs) to run through the LLM (note: this is unsupervised learning).

How exactly the pre-training works doesn’t really matter for this post? The key points to take away here are: it takes a lot of hardware, a lot of time, a lot of money, and a lot of data. So it’s pretty common for companies like OpenAI to train these LLMs and then license out their services to people to fine-tune them for their own AI applications (more on this in the next section). Also, LLMs don’t actually “know” anything in general, but at this stage in particular, they are really just trying to mimic human language (or rather what they were trained to recognize as human language).

To help illustrate what this base LLM ‘intelligence’ looks like, there’s a thought exercise called the octopus test. In this scenario, two people (A & B) live alone on deserted islands, but can communicate with each other via text messages using a trans-oceanic cable. A hyper-intelligent octopus listens in on their conversations and after it learns A & B’s conversation patterns, it decides observation isn’t enough and cuts the line so that it can talk to A itself by impersonating B. So the thought exercise is this: At what level of conversation does A realize they’re not actually talking to B?

In theory, if A and the octopus stay in casual conversation (ie “Hi, how are you?” “Doing good! Ate some coconuts and stared at some waves, how about you?” “Nothing so exciting, but I’m about to go find some nuts.” “Sounds nice, have a good day!” “You too, talk to you tomorrow!”), there’s no reason for A to ever suspect or realize that they’re not actually talking to B because the octopus can mimic conversation perfectly and there’s no further evidence to cause suspicion.

However, what if A asks B what the weather is like on B’s island because A’s trying to determine if they should forage food today or save it for tomorrow? The octopus has zero understanding of what weather is because its never experienced it before. The octopus can only make guesses on how B might respond because it has no understanding of the context. It’s not clear yet if A would notice that they’re no longer talking to B -- maybe the octopus guesses correctly and A has no reason to believe they aren’t talking to B. Or maybe the octopus guessed wrong, but its guess wasn’t so wrong that A doesn’t reason that maybe B just doesn’t understand meteorology. Or maybe the octopus’s guess was so wrong that there was no way for A not to realize they’re no longer talking to B.

Another proposed scenario is that A’s found some delicious coconuts on their island and decide they want to share some with B, so A decides to build a catapult to send some coconuts to B. But when A tries to share their plans with B and ask for B’s opinions, the octopus can’t respond. This is a knowledge-intensive task -- even if the octopus understood what a catapult was, it’s also missing knowledge of B’s island and suggestions on things like where to aim. The octopus can avoid A’s questions or respond with total nonsense, but in either scenario, A realizes that they are no longer talking to B because the octopus doesn’t understand enough to simulate B’s response.

There are other scenarios in this thought exercise, but those cover three bases for LLM ‘intelligence’ pretty well: they can mimic general writing patterns pretty well, they can kind of handle very basic knowledge tasks, and they are very bad at knowledge-intensive tasks.

Now, as a note, the octopus test is not intended to be a measure of how the octopus fools A or any measure of ‘intelligence’ in the octopus, but rather show what the “octopus” (the LLM) might be missing in its inputs to provide good responses. Which brings us to the final 1% of training, the fine-tuning stages;

LLM Interfaces

As mentioned previously, LLMs only mimic language and have some key issues that need to be addressed:

LLM base models don’t like to answer questions nor do it well.

LLMs have token limitations. There’s a limit to how much input they can take in vs how long of a response they can return.

LLMs have no memory. They cannot retain the context or history of a conversation on their own.

LLMs are very bad at knowledge-intensive tasks. They need extra context and input to manage these.

However, there’s a limit to how much you can train a LLM. The specifics behind this don’t really matter so uh… *handwaves* very generally, it’s a matter of diminishing returns. You can get close to the end goal but you can never actually reach it, and you hit a point where you’re putting in a lot of work for little to no change. There’s also some other issues that pop up with too much training, but we don’t need to get into those.

You can still further refine models from the pre-training stage to overcome these inherent issues in LLM base models -- Vicuna-13b is an example of this (I think? Pretty sure? Someone fact check me on this lol).

(Vicuna-13b, side-note, is an open source chatbot model that was fine-tuned from the LLaMA model using conversation data from ShareGPT. It was developed by LMSYS, a research group founded by students and professors from UC Berkeley, UCSD, and CMU. Because so much information about how models are trained and developed is closed-source, hidden, or otherwise obscured, they research LLMs and develop their models specifically to release that research for the benefit of public knowledge, learning, and understanding.)

Back to my point, you can still refine and fine-tune LLM base models directly. However, by about the time GPT-2 was released, people had realized that the base models really like to complete documents and that they’re already really good at this even without further fine-tuning. So long as they gave the model a prompt that was formatted as a ‘document’ with enough background information alongside the desired input question, the model would answer the question by ‘finishing’ the document. This opened up an entire new branch in LLM development where instead of trying to coach the LLMs into performing tasks that weren’t native to their capabilities, they focused on ways to deliver information to the models in a way that took advantage of what they were already good at.

This is where LLM interfaces come in.

LLM interfaces (which I sometimes just refer to as “AI” or “AI interface” below; I’ve also seen people refer to these as “assistants”) are developed and fine-tuned for specific applications to act as a bridge between a user and a LLM and transform any query from the user into a viable input prompt for the LLM. Examples of these would be OpenAI’s ChatGPT and Google’s Bard. One of the key benefits to developing an AI interface is their adaptability, as rather than needing to restart the fine-tuning process for a LLM with every base update, an AI interface fine-tuned for one LLM engine can be refitted to an updated version or even a new LLM engine with minimal to no additional work. Take ChatGPT as an example -- when GPT-4 was released, OpenAI didn’t have to train or develop a new chat bot model fine-tuned specifically from GPT-4. They just ‘plugged in’ the already fine-tuned ChatGPT interface to the new GPT model. Even now, ChatGPT can submit prompts to either the GPT-3.5 or GPT-4 LLM engines depending on the user’s payment plan, rather than being two separate chat bots.

As I mentioned previously, LLMs have some inherent problems such as token limitations, no memory, and the inability to handle knowledge-intensive tasks. However, an input prompt that includes conversation history, extra context relevant to the user’s query, and instructions on how to deliver the response will result in a good quality response from the base LLM model. This is what I mean when I say an interface transforms a user’s query into a viable prompt -- rather than the user having to come up with all this extra info and formatting it into a proper document for the LLM to complete, the AI interface handles those responsibilities.

How exactly these interfaces do that varies from application to application. It really depends on what type of task the developers are trying to fine-tune the application for. There’s also a host of APIs that can be incorporated into these interfaces to customize user experience (such as APIs that identify inappropriate content and kill a user’s query, to APIs that allow users to speak a command or upload image prompts, stuff like that). However, some tasks are pretty consistent across each application, so let’s talk about a few of those:

Token management

As I said earlier, each LLM has a token limit per interaction and this token limitation includes both the input query and the output response.

The input prompt an interface delivers to a LLM can include a lot of things: the user’s query (obviously), but also extra information relevant to the query, conversation history, instructions on how to deliver its response (such as the tone, style, or ‘persona’ of the response), etc. How much extra information the interface pulls to include in the input prompt depends on the desired length of an output response and what sort of information pulled for the input prompt is prioritized by the application varies depending on what task it was developed for. (For example, a chatbot application would likely allocate more tokens to conversation history and output response length as compared to a program like Sudowrite* which probably prioritizes additional (context) content from the document over previous suggestions and the lengths of the output responses are much more restrained.)

(*Sudowrite is…kind of weird in how they list their program information. I’m 97% sure it’s a writer assistant interface that keys into the GPT series, but uhh…I might be wrong? Please don’t hold it against me if I am lol.)

Anyways, how the interface allocates tokens is generally determined by trial-and-error depending on what sort of end application the developer is aiming for and the token limit(s) their LLM engine(s) have.

tl;dr -- all LLMs have interaction token limits, the AI manages them so the user doesn’t have to.

Simulating short-term memory

LLMs have no memory. As far as they figure, every new query is a brand new start. So if you want to build on previous prompts and responses, you have to deliver the previous conversation to the LLM along with your new prompt.

AI interfaces do this for you by managing what’s called a ‘context window’. A context window is the amount of previous conversation history it saves and passes on to the LLM with a new query. How long a context window is and how it’s managed varies from application to application. Different token limits between different LLMs is the biggest restriction for how many tokens an AI can allocate to the context window. The most basic way of managing a context window is discarding context over the token limit on a first in, first out basis. However, some applications also have ways of stripping out extraneous parts of the context window to condense the conversation history, which lets them simulate a longer context window even if the amount of allocated tokens hasn’t changed.

Augmented context retrieval

Remember how I said earlier that LLMs are really bad at knowledge-intensive tasks? Augmented context retrieval is how people “inject knowledge” into LLMs.

Very basically, the user submits a query to the AI. The AI identifies keywords in that query, then runs those keywords through a secondary knowledge corpus and pulls up additional information relevant to those keywords, then delivers that information along with the user’s query as an input prompt to the LLM. The LLM can then process this extra info with the prompt and deliver a more useful/reliable response.

Also, very importantly: “knowledge-intensive” does not refer to higher level or complex thinking. Knowledge-intensive refers to something that requires a lot of background knowledge or context. Here’s an analogy for how LLMs handle knowledge-intensive tasks:

A friend tells you about a book you haven’t read, then you try to write a synopsis of it based on just what your friend told you about that book (see: every high school literature class). You’re most likely going to struggle to write that summary based solely on what your friend told you, because you don’t actually know what the book is about.

This is an example of a knowledge intensive task: to write a good summary on a book, you need to have actually read the book. In this analogy, augmented context retrieval would be the equivalent of you reading a few book reports and the wikipedia page for the book before writing the summary -- you still don’t know the book, but you have some good sources to reference to help you write a summary for it anyways.

This is also why it’s important to fact check a LLM’s responses, no matter how much the developers have fine-tuned their accuracy.

(*Sidenote, while AI does save previous conversation responses and use those to fine-tune models or sometimes even deliver as a part of a future input query, that’s not…really augmented context retrieval? The secondary knowledge corpus used for augmented context retrieval is…not exactly static, you can update and add to the knowledge corpus, but it’s a relatively fixed set of curated and verified data. The retrieval process for saved past responses isn’t dissimilar to augmented context retrieval, but it’s typically stored and handled separately.)

So, those are a few tasks LLM interfaces can manage to improve LLM responses and user experience. There’s other things they can manage or incorporate into their framework, this is by no means an exhaustive or even thorough list of what they can do. But moving on, let’s talk about ways to fine-tune AI. The exact hows aren't super necessary for our purposes, so very briefly;

Supervised fine-tuning

As a quick reminder, supervised learning means that the training data is labeled. In the case for this stage, the AI is given data with inputs that have specific outputs. The goal here is to coach the AI into delivering responses in specific ways to a specific degree of quality. When the AI starts recognizing the patterns in the training data, it can apply those patterns to future user inputs (AI is really good at pattern recognition, so this is taking advantage of that skill to apply it to native tasks AI is not as good at handling).

As a note, some models stop their training here (for example, Vicuna-13b stopped its training here). However there’s another two steps people can take to refine AI even further (as a note, they are listed separately but they go hand-in-hand);

Reward modeling

To improve the quality of LLM responses, people develop reward models to encourage the AIs to seek higher quality responses and avoid low quality responses during reinforcement learning. This explanation makes the AI sound like it’s a dog being trained with treats -- it’s not like that, don’t fall into AI anthropomorphism. Rating values just are applied to LLM responses and the AI is coded to try to get a high score for future responses.

For a very basic overview of reward modeling: given a specific set of data, the LLM generates a bunch of responses that are then given quality ratings by humans. The AI rates all of those responses on its own as well. Then using the human labeled data as the ‘ground truth’, the developers have the AI compare its ratings to the humans’ ratings using a loss function and adjust its parameters accordingly. Given enough data and training, the AI can begin to identify patterns and rate future responses from the LLM on its own (this process is basically the same way neural networks are trained in the pre-training stage).

On its own, reward modeling is not very useful. However, it becomes very useful for the next stage;

Reinforcement learning

So, the AI now has a reward model. That model is now fixed and will no longer change. Now the AI runs a bunch of prompts and generates a bunch of responses that it then rates based on its new reward model. Pathways that led to higher rated responses are given higher weights, pathways that led to lower rated responses are minimized. Again, I’m kind of breezing through the explanation for this because the exact how doesn’t really matter, but this is another way AI is coached to deliver certain types of responses.

You might’ve heard of the term reinforcement learning from human feedback (or RLHF for short) in regards to reward modeling and reinforcement learning because this is how ChatGPT developed its reward model. Users rated the AI’s responses and (after going through a group of moderators to check for outliers, trolls, and relevancy), these ratings were saved as the ‘ground truth’ data for the AI to adjust its own response ratings to. Part of why this made the news is because this method of developing reward model data worked way better than people expected it to. One of the key benefits was that even beyond checking for knowledge accuracy, this also helped fine-tune how that knowledge is delivered (ie two responses can contain the same information, but one could still be rated over another based on its wording).

As a quick side note, this stage can also be very prone to human bias. For example, the researchers rating ChatGPT’s responses favored lengthier explanations, so ChatGPT is now biased to delivering lengthier responses to queries. Just something to keep in mind.

So, something that’s really important to understand from these fine-tuning stages and for AI in general is how much of the AI’s capabilities are human regulated and monitored. AI is not continuously learning. The models are pre-trained to mimic human language patterns based on a set chunk of data and that learning stops after the pre-training stage is completed and the model is released. Any data incorporated during the fine-tuning stages for AI is humans guiding and coaching it to deliver preferred responses. A finished reward model is just as static as a LLM and its human biases echo through the reinforced learning stage.

People tend to assume that if something is human-like, it must be due to deeper human reasoning. But this AI anthropomorphism is…really bad. Consequences range from the term “AI hallucination” (which is defined as “when the AI says something false but thinks it is true,” except that is an absolute bullshit concept because AI doesn’t know what truth is), all the way to the (usually highly underpaid) human labor maintaining the “human-like” aspects of AI getting ignored and swept under the rug of anthropomorphization. I’m trying not to get into my personal opinions here so I’ll leave this at that, but if there’s any one thing I want people to take away from this monster of a post, it’s that AI’s “human” behavior is not only simulated but very much maintained by humans.

Anyways, to close this section out: The more you fine-tune an AI, the more narrow and specific it becomes in its application. It can still be very versatile in its use, but they are still developed for very specific tasks, and you need to keep that in mind if/when you choose to use it (I’ll return to this point in the final section).

[pearlywritings]

Token of appreciation event

I deeply appreciate all the love and support my followers show me, so this event I decided to dedicate to this feeling. And nothing, in my opinion, could help me better than the amazing manga “Veil”. I highly recommend it for reading and, using some of the citations, do not claim any of its contents - all rights belong to the author.

Masterlist

Rules:

character x fem!reader. Plus please remember that I prefer writing human x human, immortal x immortal etc.

1 character if you want a ~1k words drabble and up to 3 characters if you want snippets.

1-2 prompts per request (2 prompts are only for 1 character cases).

It can be written either in a canon setting or in a modern one

smut, angst, fluff - anything

will be written with the established relationship in mind, so please state specifically if you imply another kind of relationship, or you can specify, if you want the characters to be married or still dating.

you CAN change some words within the prompt. But if you send only a number - I am following the original one.

Characters I can or can try to write about: 

Albedo, Alhaitham, Capitano, Childe, Crepus, Dainsleif, Diluc, Kaeya, Ayato, Kaveh, Lyney, Neuvillette, Pierro, Thoma, Tighnari, Wriothesley, Zhongli; Blade, Gepard, Jing Yuan, Loucha, Sampo.

Your request, if ALL things are considered, will look something like this:

Character’s name(s), prompt(s) in number or in text if it’s altered, canon/modern setting, genre, relationship status (optional) + you can add some details that I am free to both utilize or decline.

But it also can be shorter if you want to give me more freedom!

I hope we all will have fun during this event ^^

Prompts:

1)

- You can’t go a minute without me, can you?

- No, i can’t go a minute without you

2) 

- …Sometimes it's hard to tell whether you're serious or just joking.

- That's actually my special skill.

3) 

- How is it?

- Really good.

- My word, is that the only thing you can say?

- It really is beautiful. Ah… with those rosy cheeks, it's stunning. Madam…

- Puh! Surely, you jest.

4) 

- What will we be doing today?

- I have an idea - sleep in.

5)

- Did we have something planned today?

- No, nothing. In fact… let's think about it in bed.

6) 

- By the way… If I like dressing up so much, it's only your fault!

7)

- Well, well… You shouldn't stay like that, a thousand steps away from me. Stay close to me, alright?

8)

- Your boots make a very strong and imposing sound.

- My boots are honored to be admired, however they also recommend you not to stray too far.

9)

- You don't say "enter!" without at least putting a bathrobe on. Or anything at all!

- I didn't want to keep you waiting.

10)

- Oh… you mean you'll write your name? On me? Oh you!

11)

- Say give me your hand? Hm… I see.

- What's going on?

- Apparently the size of your hand is the same size as your heart.

12)

- You know I like the face you make when you smell something you like.

- I'll start charging an exhibit fee then…

13)

- Are you trying to tell me I should carry you to bed?

- My room is soooo far away, at least twenty steps from here… I'm afraid you won't be able to carry me that far?

- My room is right here.

14)

- I dreamt I was touching your hair.

- How was it?

- Well… I forgot…

- Oh, the thing is… I don't let just anyone touch my hair.

15)

- Helloooooo?

- It's so nice of you to wake me up before my alarm does…

- Your alarm has the right to rest on Sundays. At least I think so…

16)

- You… You are not going to ask me how I think you look?

- What do you think?

- You're oh so handsome.

17)

- I'm often told I have a mean glare…

- Definitely not. Anywho says that has never truly looked at you.

18)

- The razor is sharp. 

- They are made to be dangerous. But I know I don't need anything sharp to make a braid.

- You'd like to braid my hair? I can show you!

19)

- I am wondering what could be so distracting that you couldn't hear my voice anymore.

- Your hands.

20)

- It… it's the first time I've heard them described that way.

- It's the first time I used such words too.

21)

- …and above all, do not let go… For any reason, understood?

- If you tell me this one more time, I will let go.

22)

- I'm being serious right here!

- If you are so worried about it, just handcuff me to you!

23)

- Ah! Your shoelace is undone! On the left…

- I can tie it on my own!

- Well… Too late, I already did it.

24)

- I'm the one keeping a spare key to your place?

- Of course. What am I supposed to do with it? I already have one.

25)

- Saying my hair looks like glass, and my hands are flowers… how romantic you are. Or are they just words from the book?

- Only if I were to write them down.

26)

- That free cigarette looks so inviting.

- Nuh-uh!

- Ooh, so scary. Is that your last one?

- There's lipstick on it… I'll have her smoke it…

27)

- Is your bed cramped when we sleep together?

- It is. In the good way.

[iamqueenpotato]

What We Need

Lucien x Reader

A/N- So I have been obsessed with the band Sleep Token for the past month or so and while listening to their song Granite, this little fic popped into my head. Again I appreciate you all so much, it makes my day when I see you all enjoy my work! Stay amazing 💜

Summary: Y/N is distraught after a fight with Lucien. But she decides to visit him. To finally speak the truth about how she feels.

Enjoy!

⚠️: None

Word Count: 1k

You shouldn’t have gone back to him, you knew it was wrong. He had a mate. And the two of you weren’t exactly on the best terms with each other either. But after your fight earlier in the evening it drove a wedge between the two of you. One that you thought would be irreversible. Yet you couldn’t blame him, he had found his mate, and though she did not show interest he still ended things between the two of you. Claiming the bond would make things difficult, and that you and he shouldn’t talk anymore, saying it wouldn't benefit either of you.

You wouldn’t lie and say it wasn't painful hearing those words, Lucien meant so much to you and the way he spoke was as if the relationship you had shared meant nothing to him. So you let the anger get the best of you, and maybe he was just a fling for the time or maybe it was something else. Something more. But most importantly he was once your best friend. And the thought of losing that terrified you.

So here you were standing in front of Lucien’s door. You let your hand float in the air, ignoring the slight shake before it came down on the door to alert him of your presence. You took a deep breath in, attempting to calm your nerves as the door opened softly, Lucien standing on the other side, his hair was tied back loosely, a few strands falling in front of his face. He was wearing nothing but loose pajama pants, ones that hung low on his hips. You brought your attention back to his face, ignoring the sudden indecent thoughts. His eyes widened when he saw you. “Y/N. What are you doing here? It’s late.” He spoke quietly, assuming he didn’t want to wake anybody that could be staying in the rooms down the hall. Though his voice was quiet, his tone wasn’t exactly welcoming. You know now this wasn’t the same Lucien you had grown up with.

“I want you but I know I am not what you need. Lucien, you were more than somebody I was destined to meet. And tonight you gave me nothing whatsoever but a reason to leave. And if this is the last night I see you then so be it, but I don’t want things to end the way they did. I cannot leave with anger between us. You’re my best friend. My first love.” You paused, unable to read his expression, and you began to doubt yourself. Maybe this wasn’t the best idea to be here. But you powered through, if he didn’t want to see you ever again after this then you would live with it. Even if it pained you. “I get your situation. I do. But I will not be a memory. I refuse to be. And I want you to know that however things work out between you and your mate I will still be your best friend. Even if you don’t want me to. I will still be here.” You let out a deep sigh and you could feel your hands shaking, it was time he knew the truth, if this was the end of the story you had shared with him, you found no harm in putting it all out there. He continued to stare at you, his hand tightly gripping the edge of the door. You pursed your lips into a thin line, accepting his silence as an answer. Turning around you started walking down the hallway. Ignoring the creeping pain in your chest. If he was happy then you would be happy. Even if that meant you were no longer a part of his life.

Suddenly, you felt someone grab your arm, turning your body to face them. Lucien held on to your arm tightly, but it didn’t hurt. You watched him as he stared at you, he hadn’t said a word, and the two of you stood in silence until you decided to break it.

“Lucien, what are you-” Your words were cut short when he pulled you close to him, crashing his lips into yours. Naturally, you melted into him, sliding your hands up his chest, slowly wrapping them around his neck. He snaked one arm around your waist, while the other supported the back of your head. He kissed you with that fiery passion you adored, he held you close as if he was afraid he was going to lose you at any moment. Your lips moved together, each breath shared as the two of you held each other.

You finally pulled away, looking up at him with hooded eyes. Your heavy breaths filling the silence of the empty hall. You could hear your heartbeat within your chest.“I didn’t want to leave you. I was selfish to think it would easier to be without you, but you are what I need, there is nothing or no one that could replace you. After finding out about Elain I was confused. I felt obligated to her, but Y/N, my heart belongs to you. I thought it would be easier to let it go, to make you hate me. But I was wrong. I’d rather be six feet under than live this life without you.” Lucien admitted. You didn’t know when you started crying but the tears fell slowly down your cheeks. He moved his hands to cup your face, wiping his thumbs along your cheekbones. “I am so sorry my love. Stay with me. Please.” He whispered, resting his forehead on yours. Closing your eyes you nodded, lifting your head to kiss him once more. He smiled as his lips lingered on yours, moving his hands to interlace his fingers with your own. Lucien pulled away slightly. “My heart is yours. Always has been.” His words were soft as they brushed against your lips. And a smile crept across your face.

“And mine is yours.” You whispered, and the fear of losing him and that relationship you cherished soon became a distant memory.

Taglist: Taglist: @honestlywtfisgoingon @fanfictioniseverything @marina468 @positivewitch @maviee @blurredlamplight @bookslut420 @starswholistenanddreamsanswered @elle10 @dragonstoneprincess

(If you guys don't want to be tagged in Lucien fics let me know! I'm probably going to start a separate tag list for Lucien and Azriel ) 💜

[thebusytypewriter]

Salem's Holiday Season Song Event

Hello hello! Thank you for the fantastic reception of the previous event! You guys really brought some fantastic requests. We’ve grown quite a bit as a result, and I couldn’t be more grateful! As another token of gratitude, it’s time for an event! But this time will be different… Let me explain.

I’m opening requests again, but with a twist: you can send me a character or ship, along with a number 1-300, and I will write a ficlet (≤1k words)! The number corresponds to a song on my Liked Songs playlist from Spotify, and that will be the basis for your fic.

HOWEVER! I will only be taking one request per number. This is a first come, first serve choice, and I will be updating this post with every request. If you request a number that is already chosen, I will either answer the request (privately if not anon) and ask for a second choice. It won’t move your place in the queue, don’t worry!

(I will, however, do multiple fics for a character/ship/etc. so you don’t need to worry about that, either :>)

This is for anyone and everyone—follower or not! (Though following is also appreciated!) All requests are to be sent via ask box, please and thank you!

The FAQ/rules are listed below.

What can I request? I accept requests for any sort of fic, including but not limited to: reader insert, canon characters, ships, etc!

What fandoms are up for request? Any fandom I have posted for is up for grabs. The most notable fandoms include Danganronpa, Gravity Falls, Steven Universe, Over the Garden Wall, Fullmetal Alchemist, Undertale/Deltarune, Portal / Portal 2, The Adventure Zone, The Owl House, The Dark Crystal, The Amazing Digital Circus, and InuYasha. This does not exclude anything not listed here, so get creative!

What are your boundaries? I have few boundaries, but they are significant! Please do not request anything involving serious gore, smut, pedophilia, incest, or anything close to those categories. If I deem your request outside of my boundaries, I will either a) respond privately to alert you and request an alternative if desired or b) delete it if it is on anon (as I believe I cannot respond privately if you are on anon).

Where can I get an idea of your writing? If you want to see how I write for consideration purposes, you can find my writing masterlist on the pinned About Me post, or you can snoop around my AO3.

Requests will be open for—tentatively—four weeks (closing approximately 12/8, 11:59pm CST), then I will be closing the ask box to work on them. So get that ask in there before it’s too late!

Currently Unavailable: 1, 3, 5, 10, 13, 53, 69, 162, 283, 300

And I once again will be publishing each request on AO3 in a designated series, just so you're aware! :D

Thank you guys again for your support, and happy requesting!

- Salem (thebusytypewriter)

[thedevilspearl]

🪷 ᴍɪʟᴇsᴛᴏɴᴇ ᴍᴏᴏᴅʙᴏᴀʀᴅs 🪷

so i hit 1k followers and i just—! it’s really amazing and i can’t believe this blog has grown this much when it hasn’t even been six months since i started >_< thank you so much!!!!

i haven’t done an event for so long so i decided to do one for this milestone! it’s a moodboard event for you and your fave or your self ship for obey me characters only.

🪷 to take part, you must be following me since this is a follower event. if you follow from another blog, let me know

🪷 it is not a requirement but it would be courteous to reblog my masterlist not only to help promote my works but it will act as a token of appreciation for me (ie. i will feel more motivated if you actively show your appreciation)

🪷 anons are allowed and i will assume all the other guidelines are followed, but if i get a blunt and demanding request, especially on anon, i will not do that request

now onto what you need to send me:

🪷 firstly, tell me which of my works is your favourite and why?

➛ it sounds silly, but if you’re a follower, you most likely have read my writing and i wanna know which ones you like the most and why you love them so much <3

🪷 then tell me your character of choice. it can be your fave or a character you self ship with.

➛ if necessary, please tell me the gender coupling. my default will be m/f if you don’t specify otherwise. you can also request for me to avoid using images of people/couples and focus on depicting the vibes only.

🪷 lastly, tell me five things which describe or relate to your relationship with the character. keep them short and sweet!

➛ make sure they are things (like items, places or actions) and not you just describing what the ship is like. eg. books, holding hands under the table, summer rain, the colour blue, sharing clothes, expensive sports cars....

you get what i mean. i don’t want to sift through five paragraphs of you telling me about the ship. just give me a list of five things, quotes and songs are also okay!

[emeritus-fuckers]

Congrats on 1k followers! I would love to take part in the match up rerun if possible.

1. He/him/they/them pronouns, transmasculine pansexual!

2. Either Papas or ghouls, I love everyone.

3. I'm 5'3" and on the chubby side, but I do have a fair amount of muscle underneath the chub. My hair is shaved around the sides and back, but on top I've got long curly hair that is past my shoulders and lavender colored. My clothing style I kind of refer to as comfort goth grunge. Usually always black and pops of color, but generally you'll find me in distressed oversized sweaters or hoodies with bleach dyed sweat pants.

4. My friends describe me as chaotic but loving. Very much golden retriever kind of energy. I have a strong sense of empathy, and so a lot of people often confide in me or seek my help for personal issues.

5. I love writing, painting, drawing, reading, and playing video games. At the moment my favorite bands are Ghost, Sleep Token, and the Misfits. Hannibal and RuPaul's Drag Race are the two shows I watch on repeat. I also am kind of obsessed with stuffed animals and bones. There are probably over 100 stuffed animals in my bedroom, and have one shelf that is full of assorted animal skulls I've either found in nature or have been gifted by friends(I name each one).

6. I grew up with a twin, so I am very used to always having someone around. So I can easily get separation anxiety if left home alone for more than a couple of days. Doesn't even need to be a specific person, I just am comforted by having another individual within close proximity.

Thank you so much! 💙

This post is part of the 1000 followers match up event. Entries for the event are now closed.

Your match is...Papa Emeritus III

You two bonded over having a similar energy.

Terzo is very sociable and likes to be around people so you two started hanging out more and more. He found himself thinking about you more and more so he would always seek you out.

Then he asked you on a date to go to a painting class with him.

He spent the entire time amazed at how good your work is. Terzo is very enthusiastic but he isn’t the best technically at painting. Still there is an energy there in his work, and the grin he gives you as he shows you is just adorable.

So adorable that you give him a kiss on the cheek to say well done.

If you are comfortable with it, he’ll sweep you up in his arms twirling you around and then kissing you.

He loves sitting on the sofa with you in his arms while you read.

Sometimes it’s while you try to read as Terzo is always trying to get your attention. But in a very sweet way. He also knows when to stop and actually let you read. He’s started reading too.

He loves it when you both finish a book and then talk about it late into the night

~

Written by Nyx

[sundayevenings11pm]

31th Oct Monday - Lei

Recognizing Progress 

I haven’t written here for quite a while but let me just recognize the progress I’ve made for the last two weeks. 

The first big thing that motivated me was that I wanted to be a participant for the evmos hackathon. The cosmos ecosystem is really interesting and if I could win some prizes at the hackathon, that would be an even greater win. 1k worth of evmos tokens might just be worth 100k during the next bull run. So I shouldn’t underestimate the amount even a small winning could get me. 

To get to the stage where I could produce something good, I need to be good at typescript. There are many projects that’s written in typescript, and quite frankly, if I become a typescript ninja, I have one of the most valuable coding skills of today. Typescript is tough, but it is not that tough. It’ll enable me to code very robust apps, that’ll drastically increase my skillset as a blockchain developer. 

So last week, I spent the entire week working more than 3 hours a day on Week 9 of the fullstackopen course on typescript. It is tedious and at times frustrating, but fulfilling nonetheless. 

In summary: I spent much of my time 2 weeks ago on the Evmos ecosystem, and last week finessing my typescript. 

This week started kind of tough because I realized that the fullstackopen course got to a point where it’s way more difficult than what I’ve expected. I am dealing with a couple of more ‘advanced’ concepts such as react hooks like useContext and useReducer - the former of which I have just been exposed to, and the latter of which is pretty rusty to me. On top of that, i am implementing things like types to each of those. So it started to get confusing very quickly. 

I stepped back and evaluated the estimated time I’ll take to finish just this part of the course, which is around a week. Looking at the number of things on my table, I felt that I needed to perhaps readjust my focus back to some work elements again. The big hackathon starts next week, and I’ll need to begin ensuring that all the pieces are in place. On top of that, I’ll need to dive deeper into the evmos ecosystem and see what else I can learn there. 

This week is also the start of sprint #1 of Q4. There are a couple of pieces on my table for the next two weeks - should be a fun and productive two weeks. 

Goals for the week 

Work:

- Ensure that hackathon matters get looked into (everyday 1 hr)

- Channel sales support (1hr)

- Identify large sponsors for Fantom Hackathon (1hr)

- Workshops (EthDenver) - focus on perfecting a version (2hrs)

- Review Submissions for Encode x Harmony (2hrs)

- Begin on Dex Pools table component (review by EoW) (2 x 3hrs)

Quite a lot of stuff for work already this week. 

Learning: 

- Slowly learn and fully master 9d of Typescript course so that I can start using it in my build for evmos. (3 x 3hrs)

- Evmos ecosystem stuff (5 hrs)

Personal Projects:

- Mum’s website (3hrs)

Let’s go. Bali in two weeks. 

[sdyuteiaok]

Drafted MH3, and had a pretty poor showing. The deck really didn't have much juice, but I feel like I could have scrapped together at least one or two more wins, but we could only manage two. I also had a couple poor picks as I was reading cards and evaluating on the fly--I should have had a second copy of the -3/-3, +1 counter trick, and I could have bolstered my lands with the UG land to both help my splash, and have a monster on the top end. I was BG mainly, but had the target-twice legendary bird as my U splash. So that dual land would have been a welcomed addition, but I discounted it because I shortcutted it as purely a U card, but I coulda cast the G side. One game felt like it was in reach, but I just couldn't get anything to push the issue when he had a big flying blocker, and with 6 cards left, he cast the Gremlins card with 14 energy so, even though I could trade on most things, he had enough Spawn to close it out after some early pressure had diminished my life enough to make the swing worthwhile. But I was bleeding the guy out with my deathtouch Kamigawa guy, and then eventually plinking away with my deathtouch snake from my Ophidiomancer, which I could sac to my vehicle, even. But I just needed to draw some removal or something to break through in the air, or my hydra to make my flying threats serious damage. I had him in just a couple turns. Man, if I had drawn my bestow Hydra, I bet that woulda made the difference because it gives trample, crucially, and I had a ton of mana to work with. I wonder if I shoulda been attacking more liberally since I had my Brushwagg giving most of my guys trample, so either he'd have to trade down his creature count, making the eventual Gremlin card considerably less relevant, and otherwise, he'd be taking a lot of pain every turn, restricting his moves. And then if I'd drawn my UG bird earlier, I coulda have started going off with targeting stuff, but I drew it so late I didn't get a chance to pop off with it. Pretty frustrating. I am almost positive I could have scratched that one out, but who knows. So, looking at the gem totals, I can get both passes, so maybe I should just get the MH3 one so I can start recuperating gold for the next draft, and then in the meantime, I'll be able to buy the draft token with tickets. Or maybe I should just ignore it for now, since there's still a lot of days left on the pass, and just use gems for a draft. But I currently have only about 520 gems disposable, meaning I can't afford both passes and a draft, I'm about 1k short...surely I can get that back, though...hmm.