INDEX

Explanations

symbols and punctuation related to code or mathematical expressions

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

code preprocessor directives

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

ReusableCell

-0.61

 classNames

-0.46

classnames

-0.44

 ModelExpression

-0.44

TagMode

-0.40

 bağlantılar

-0.39

 arşivlendi

-0.38

Дереккөздер

-0.38

exitRule

-0.37

StoryboardSegue

-0.36

POSITIVE LOGITS

хьтан

0.70

səhifə

0.66

FunctionFlags

0.53

RTGC

0.52

niająca

0.47

NameIn

0.44

География

0.43

'){

0.42

Impl

0.41

}{%

0.41

Activations Density 0.003%

symbols and punctuation related to code or mathematical expressions

code preprocessor directives

No Comments

No Known Activations