INDEX

Explanations

code comments and documentation format indicators

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

start of user input

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

 queſta

-0.90

 Administrativna

-0.83

 protoimpl

-0.78

writeFieldEnd

-0.77

EndContext

-0.71

ロウィン

-0.70

 surla

-0.68

oredCriteria

-0.68

 <>",

-0.68

帖最后由

-0.68

POSITIVE LOGITS

 Brust

0.35

Voci

0.31

package

0.31

참고

0.30

gebung

0.29

Geografía

0.29

 tiếp

0.28

 altas

0.27

wertes

0.27

 actually

0.27

Activations Density 0.018%

code comments and documentation format indicators

start of user input

No Comments

No Known Activations