INDEX

Explanations

mentions of specific individuals or identifiers related to personal history or characteristics

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

two-character compounds

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

parsedMessage

-0.69

SharedDtor

-0.62

featureID

-0.59

WireFormatLite

-0.58

Билгалдахарш

-0.55

 estekak

-0.55

хьтан

-0.55

 betweenstory

-0.55

 Numerade

-0.54

ſelf

-0.54

POSITIVE LOGITS

the

0.57

0.49

0.44

 ال

0.43

 geçen

0.42

 keduanya

0.42

 soprav

0.41

 mögliche

0.40

two

0.40

Activations Density 0.019%

mentions of specific individuals or identifiers related to personal history or characteristics

two-character compounds

No Comments

No Known Activations

mentions of specific individuals or identifiers related to personal history or characteristics

two-character compounds

No Comments

No Known Activations