INDEX

Explanations

references to honors or awards given to individuals

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

honorary titles and roles

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

migrationBuilder

-0.71

tagHelperRunner

-0.66

 Theſe

-0.65

 Monfieur

-0.63

ագրություններ

-0.59

 Anſ

-0.59

tvguidetime

-0.58

 مرئيه

-0.58

 informée

-0.58

.*")]

-0.56

POSITIVE LOGITS

 honorary

0.54

 Honorary

0.47

Honorary

0.40

GREES

0.31

 graduation

0.30

 diploma

0.30

 humanitarian

0.30

scroll

0.29

 års

0.29

 degree

0.28

Activations Density 0.007%

references to honors or awards given to individuals

honorary titles and roles

No Comments

No Known Activations