INDEX

Explanations

mentions of years and scientific studies or reports

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

study, research, and publication context

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

featureID

-0.50

 FERR

-0.39

AntiForgeryToken

-0.39

 Ferdin

-0.39

[:

-0.39

 RUSSELL

-0.39

Ferr

-0.38

inton

-0.37

."—

-0.37

<--

-0.36

POSITIVE LOGITS

 للمعارف

0.71

ⓧ

0.60

 anonyme

0.51

 שוליים

0.48

 թվական

0.48

 חיצוניים

0.47

AndEndTag

0.46

CppMethod

0.45

 怎样

0.42

onOptions

0.40

Activations Density 0.061%

mentions of years and scientific studies or reports

study, research, and publication context

No Comments

No Known Activations