INDEX

Explanations

references to individuals' identities and background information related to legal circumstances

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

legal and investigative reports

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

RenderAtEndOf

-0.79

parsedMessage

-0.70

 typelib

-0.64

 صوتيه

-0.64

NewUrlParser

-0.64

Демографія

-0.64

 esternos

-0.63

 ModelExpression

-0.63

 ſind

-0.63

 Réponses

-0.63

POSITIVE LOGITS

 reportedly

0.32

Investigators

0.28

 arrep

0.28

raisemb

0.28

 conoc

0.28

 ebenfalls

0.27

 abond

0.27

 informaron

0.27

 apparently

0.27

 предвари

0.26

Activations Density 0.041%

references to individuals' identities and background information related to legal circumstances

legal and investigative reports

No Comments

No Known Activations