INDEX

Explanations

references to statistical or numeric data related to research findings

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

columns separated by dashes

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

antium

-0.36

omod

-0.34

 кра

-0.33

 fend

-0.32

ighter

-0.32

feas

-0.31

 digs

-0.31

ohyd

-0.31

 الر

-0.30

 JOUR

-0.30

POSITIVE LOGITS

 autorytatywna

0.70

awtextra

0.62

 PeEnEo

0.58

 estekak

0.58

WriteAttribute

0.56

SequentialGroup

0.56

AnimationsModule

0.55

 betrek

0.55

endmodule

0.54

setupUi

0.54

Activations Density 0.140%

references to statistical or numeric data related to research findings

columns separated by dashes

No Comments

No Known Activations