INDEX

Explanations

references to visual aids such as illustrations, pictures, and tables in the text

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

figures, tables, or lists

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

abestanden

-0.69

 الحره

-0.60

AndEndTag

-0.57

expandindo

-0.57

 ब्रेकडाउन

-0.54

Portale

-0.54

VERTISEMENT

-0.51

__).

-0.50

msford

-0.47

@"/

-0.46

POSITIVE LOGITS

EDEFAULT

0.44

endregion

0.43

 istrinya

0.38

 leitor

0.37

 voren

0.35

LookAnd

0.35

requestBody

0.34

 lisäksi

0.33

 gegebenen

0.33

brengen

0.33

Activations Density 0.161%

references to visual aids such as illustrations, pictures, and tables in the text

figures, tables, or lists

No Comments

No Known Activations