INDEX

Explanations

unique characters or symbols in coding or programming contexts

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

short consonant strings

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

󠁮

-0.59

UnusedPrivate

-0.56

曖昧さ回避

-0.54

 propOrder

-0.53

 EconPapers

-0.53

 Chwiliwch

-0.52

KommentareTeilen

-0.51

}{*}{}

-0.51

forChild

-0.49

})$}

-0.49

POSITIVE LOGITS

etc

0.44

rmtree

0.40

 defaultstate

0.39

etc

0.39

asdf

0.37

omento

0.36

 insanely

0.35

ug

0.34

setHeight

0.34

Activations Density 0.005%

unique characters or symbols in coding or programming contexts

short consonant strings

No Comments

No Known Activations