INDEX

Explanations

specific identifiers or markers in structured data formats

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

field list or array

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

 conmigo

-0.37

 cremallera

-0.36

accompagnement

-0.35

 industriel

-0.35

 konus

-0.34

 dudas

-0.34

 Vermögen

-0.34

 actualizaciones

-0.34

Formazione

-0.34

 kelapa

-0.33

POSITIVE LOGITS

出版年

0.84

RenderAtEndOf

0.79

]=>

0.79

 للاسماء

0.77

ScopeManager

0.77

 يتيمه

0.75

GTCX

0.73

webElement

0.73

]")]

0.72

windowFixed

0.71

Activations Density 0.097%

specific identifiers or markers in structured data formats

field list or array

No Comments

No Known Activations