INDEX

Explanations

technical terms and identifiers in code or scripts

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

block or org followed by punctuation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

fromnode

-0.56

 jScrollPane

-0.50

 للمعارف

-0.49

 defaultstate

-0.44

 صوتيه

-0.44

DropColumn

-0.44

 tartalomajánló

-0.43

BagConstraints

-0.43

料

-0.42

zet

-0.42

POSITIVE LOGITS

zbęd

0.46

 publicitaires

0.42

SharedCtor

0.41

 protoimpl

0.40

 Hauteur

0.40

RenderAtEndOf

0.38

 dimentic

0.38

 aikaa

0.38

 cenar

0.38

agré

0.37

Activations Density 0.119%

technical terms and identifiers in code or scripts

block or org followed by punctuation

No Comments

No Known Activations