the birb

data monkey by day vn dev by night | i use bad english to tell bad stories about moderately okay things | she/they | h+

[feeney ~comms open~ @feeneyonfire]

Commission info ヽ(*・ω・)ﾉ

– Payment via Hipolink
– Full prepay
– I’ll finish commission within 3 weeks after I start working on it

Send me a message if you are interested!

My other links:
https://t.co/aiNNnBTWJR
https://t.co/3spgVPGCgp
https://t.co/EQDb5BDKPm

[academic mordochka @amordochka]

do you know that my silly little yuri game is on winter sale right now?
if you hate 4th walls (as is, love breaking them) and love ladies bullying you - it might be just the thing for you
https://t.co/vV7MQPJUFQ

[academic mordochka @amordochka]

Anything can be used as a reference if you're bold enough

[anshuman @athleticKoder]

She dumped me last night.

Not because I don't listen.
Not because I'm always on my phone.
Not even because I forgot our anniversary (twice).

But because,

in her exact words:

"You only pay attention to the parts of what I say that you think are important."

I stared at her for a moment and realized...

She just perfectly described the attention mechanism in transformers.

Turns out I wasn't being a bad boyfriend. I was being mathematically optimal.

See, in conversations (and transformers), you don't give equal weight to every word. Some words matter more for understanding context. Attention figures out exactly HOW important each word should be.

Here's the beautiful math:

Attention(Q, K, V) = softmax(QK^T / √d_k)V

Breaking it down:

Q (Query): "What am I looking for?"
K (Key): "What info is available?"
V (Value): "What is that info?"
d_k: Key dimension (for scaling)

Think library analogy:

You have a question (Query). Books have titles (Keys) and content (Values). Attention finds which books are most relevant.

Step-by-step with "The cat sat on the mat":

Step 1: Create Q, K, VEach word → three vectors via learned matrices W_Q, W_K, W_V
For "cat":

Query: "What should I attend to when processing 'cat'?"
Key: "I am 'cat'"
Value: "Here's cat info"

Step 2: Calculate scoresQK^T = how much each word should attend to others

Processing "sat"? High similarity with "cat" (cats sit) and "mat" (where sitting happens).

Step 3: Scale by √d_kPrevents dot products from getting too large, keeps softmax balanced.

Step 4: SoftmaxConverts scores to probabilities:

"cat": 0.4 (subject)
"sat": 0.3 (action)
"mat": 0.2 (location)
"on": 0.1 (preposition)
"the": 0.1 (article)

Step 5: Weight valuesMultiply each word's value by attention weight, sum up. Now "sat" knows it's most related to "cat" and "mat".

Multi-Head Magic:Transformers do this multiple times in parallel:

Head 1: Subject-verb relationships
Head 2: Spatial ("on", "in", "under")
Head 3: Temporal ("before", "after")
Head 4: Semantic similarity

Each head learns different relationship types.

Why This Changed Everything:

Before: RNNs = reading with flashlight (one word at a time, forget the beginning)

After: Attention = floodlights on entire sentence with dimmer switches
This is why ChatGPT can:

Remember 50 messages ago

Know "it" refers to something specific
Understand "bank" = money vs river based on context
The Kicker:Models learn these patterns from data alone. Nobody programmed grammar rules. It figured out language structure just by predicting next words.
Attention is how AI learned to read between the lines.

Just like my therapist helped me understand my focus patterns, maybe understanding transformers helps us see how we decide what matters.

Now if only I could implement multi-head attention in dating...

Still waiting for "scaled dot-product listening" to be invented.

[the birb @heyansire]

@sbtheflea @Opheliachatts @Skrunkle_Wunkus @Iddelf @RanpoeFrotter some traits are distributed over a number of genes, and some alloys can have equal dominance — this creates so-called incomplete or partial dominance of genetic traits. I.e. if you breed a red flower and a white flower, you can get a pink flower.

[the birb @heyansire]

@Ketsu_Ten katagaren deniers remind me of this meme

[the birb @heyansire]

@Ketsu_Ten if i'm to be honest the single interpretation i have is

thank god i'm not big into league of legends anymore

[the birb @heyansire]

@Ketsu_Ten ...it also bothers me that a bunch of porn bot accounts liked my replies jeez twitter sucks now

[the birb @heyansire]

@Ketsu_Ten however, sure, i get your point
i don't even ship ekko and ezreal together
and i mean, of course ezreal x lux in that context is also canon and 'more canon' you obviously can prefer it

but at the end of the day i was making a joke about the lack of ezreal in arcane

[the birb @heyansire]

@Ketsu_Ten otherwise basically no ships are canon unless they explicitly proclaim romantic love, kiss and marry on screen

which would be... a very narrow literal approach to storytelling that rejects the metaphorical and the subtle

[the birb @heyansire]

@Ketsu_Ten sure? context also matters tho. you and i are real people with continuous external circumstances outside the message.

if i write 'damn i love that guy' in a piece of media from character's pov it means i am adding at least some narrative ambiguity to the relationship portrayed

[the birb @heyansire]

@Ketsu_Ten well i mean it is kinda put into canon
https://t.co/wL3DUdISmM

[Anthony Bonato @Anthony_Bonato]

More proof techniques

[the birb @heyansire]

@PolishLynx thank you for playing!

[the birb @heyansire]

anyway, i am not popular and no one has written a cannibal mermaid AU based on my stuff (yet?) so i'm gonna get back to not using this goddamn website

[the birb @heyansire]

so you need to be honest with yourself?? if you're this attached and involved, don't engage with fanfiction. even if you're like, really curious.

but if you do and it hurts, it's not the fault of the author of cannibal mermaids lmao

[the birb @heyansire]

a few years ago, i was writing a script, and a then-friend of mine asked 'what if you get popular and people start headcanoning your characters into unrecognizable parodies of themselves'

to which i answered 'lol i want ao3 fic about my characters being cannibal mermaids gimme'

[the birb @heyansire]

however, i kinda get the emotions?? if you put so much of yourself into your writing (and let's be honest writing is very emotionally vulnerable of an act, you can't write thousands of words and not put yourself into it) it can feel very personal if people misrepresent it