INDEX

Explanations

mentions of user identification and document-related terms

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

informational identifiers or requests

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

 AssemblyCulture

-0.69

createClass

-0.54

NameValuePair

-0.50

 AssemblyVersion

-0.49

 voorbij

-0.48

ofire

-0.47

+#+#

-0.47

internalType

-0.47

 useCallback

-0.47

ModelSerializer

-0.46

POSITIVE LOGITS

SourceChecksum

0.38

 Exactos

0.36

 garçons

0.34

 rasch

0.33

claim

0.33

WebElementEntity

0.33

 dacia

0.33

"")

0.32

your

0.32

iPod

0.31

Activations Density 0.293%

mentions of user identification and document-related terms

informational identifiers or requests

No Comments

No Known Activations