INDEX

Explanations

statistical data and numerical information related to studies or analyses

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

numbers followed by percent

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

帖最后由

-0.46

 Dimen

-0.44

 outreach

-0.40

malade

-0.40

 jams

-0.39

 spreads

-0.38

 Salah

-0.37

odynamics

-0.37

rada

-0.37

Vib

-0.36

POSITIVE LOGITS

setVerticalGroup

0.69

sizeCache

0.57

UseVisualStyle

0.56

enterOuterAlt

0.53

EndTag

0.50

AnimationsModule

0.50

DockStyle

0.48

Nationalité

0.47

stdc

0.46

Javadoc

0.46

Activations Density 0.048%

statistical data and numerical information related to studies or analyses

numbers followed by percent

No Comments

No Known Activations