INDEX

Explanations

occurrences of the word "post"

oai_token-act-pair · gpt-4o-mini Triggered by @bot

references to posts or publications

oai_token-act-pair · gpt-4o-mini Triggered by @bot

view this post

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 purpoſe

-0.59

 referenties

-0.57

 PartialEq

-0.57

AndEndTag

-0.57

yntaxException

-0.56

effectively

-0.55

UIControlState

-0.55

omitempty

-0.55

 Administrativna

-0.55

EndGlobalSection

-0.55

POSITIVE LOGITS

 made

0.53

 post

0.52

 posted

0.51

chat

0.50

 spal

0.50

gra

0.49

simpleType

0.49

halen

0.48

prepareStatement

0.45

 shared

0.45

Activations Density 0.007%

occurrences of the word "post"

references to posts or publications

view this post

No Comments

No Known Activations