INDEX

Explanations

frequency adverbs indicating how often events occur

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

frequent actions or conditions

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ContentLoaded

-0.36

ToBounds

-0.34

principalTable

-0.33

 Lexikon

-0.33

 eget

-0.31

 gial

-0.31

瓢

-0.30

 preventable

-0.30

theo

-0.30

WID

-0.30

POSITIVE LOGITS

 Administrativna

0.71

estacks

0.68

AISSEE

0.59

elemField

0.56

parsedMessage

0.54

 typically

0.54

帖最后由

0.54

 Italijanski

0.53

TargetException

0.53

rungsseite

0.53

Activations Density 0.049%

frequency adverbs indicating how often events occur

frequent actions or conditions

No Comments

No Known Activations