INDEX

Explanations

coding syntax involving return statements and framework-related terms

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

text starting with `#` or `/*`

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

DeleteBehavior

-0.69

SourceChecksum

-0.65

énario

-0.56

انيف

-0.55

FailureListener

-0.55

 disambiguazione

-0.55

 distanciation

-0.55

曖昧さ回避

-0.54

isEmail

-0.53

AfterClass

-0.52

POSITIVE LOGITS

ThroughAttribute

0.69

ICEF

0.57

--)

0.57

migrationBuilder

0.56

("")]

0.54

0.52

Rohy

0.52

 Verſ

0.52

VIAF

0.52

enumii

0.51

Activations Density 0.005%

coding syntax involving return statements and framework-related terms

text starting with `#` or `/*`

No Comments

No Known Activations