INDEX

Explanations

references to software or technical support

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

android support

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

writeFieldEnd

-0.54

ArgsConstructor

-0.45

exitRule

-0.44

 disambiguazione

-0.41

WebMethod

-0.41

 jugado

-0.40

 TAMBÉM

-0.40

 wyda

-0.40

 viņ

-0.39

 RemoteException

-0.39

POSITIVE LOGITS

support

0.84

Support

0.67

 Support

0.63

 support

0.62

OGND

0.59

 defaultstate

0.58

SUPPORT

0.57

upport

0.55

 androidx

0.54

bootstrapcdn

0.53

Activations Density 0.000%

references to software or technical support

android support

No Comments

No Known Activations