INDEX

Explanations

terms related to health, safety, and professional practices

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

reporting and prevention policies

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

 createState

-0.38

cinia

-0.30

inSlope

-0.28

 lessor

-0.27

Vegeu

-0.27

ⓧ

-0.26

例文

-0.26

 decorated

-0.26

nextProps

-0.26

ouin

-0.26

POSITIVE LOGITS

ロウィン

0.57

ſelves

0.57

 مرئيه

0.56

 tranſ

0.56

ſelf

0.56

Prevention

0.56

MigrationBuilder

0.55

 resourceCulture

0.55

 houſe

0.54

 indígen

0.54

Activations Density 0.002%

terms related to health, safety, and professional practices

reporting and prevention policies

No Comments

No Known Activations