INDEX

Explanations

references to academic or organizational affiliations and licensing terms

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

legal disclaimers and licenses

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

 незавершена

-0.40

Chham

-0.35

الحياه

-0.31

 isNameExpr

-0.31

QueryParams

-0.27

出版年

-0.27

RTLR

-0.26

還是

-0.26

 explotación

-0.26

 invasión

-0.25

POSITIVE LOGITS

PerformLayout

0.87

GOTREF

0.83

createNewFile

0.80

 مشين

0.79

printStackTrace

0.79

 postIndex

0.74

存于互联网档案馆

0.73

 RISERVATA

0.72

httphttps

0.71

enderror

0.70

Activations Density 0.023%

references to academic or organizational affiliations and licensing terms

legal disclaimers and licenses

No Comments

No Known Activations