INDEX

Explanations

requests for licensing or information about obtaining materials

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

send you a copy

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

kup

-0.57

-0.47

出版年

-0.45

Skocz

-0.45

/***/

-0.44

 Custo

-0.44

__*/

-0.43

 CURIAM

-0.43

 Cust

-0.42

 escal

-0.42

POSITIVE LOGITS

 AspNetCore

0.64

RegistryLite

0.63

inSlope

0.57

culable

0.56

 Waray

0.56

 Ressources

0.54

ernalia

0.54

UrlResolution

0.54

jeune

0.54

ngrx

0.53

Activations Density 0.181%

requests for licensing or information about obtaining materials

send you a copy

No Comments

No Known Activations