INDEX

Explanations

sequences of whitespace or formatting characters in the text

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

HTML, code, numbers, and text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

glLoad

-0.44

EndProject

-0.39

 typelib

-0.39

тистика

-0.39

 salvajes

-0.39

HttpDelete

-0.37

 capuz

-0.37

↵↵

-0.36

而出

-0.36

 виправивши

-0.36

POSITIVE LOGITS

 snippetHide

0.52

Tembelea

0.48

 للاسماء

0.47

 demografica

0.47

 мәкал

0.45

 Irishman

0.45

vician

0.44

"])

0.44

 themſelves

0.43

 itſelf

0.43

Activations Density 0.000%

sequences of whitespace or formatting characters in the text

HTML, code, numbers, and text

No Comments

No Known Activations