INDEX

Explanations

terms related to job candidates and applicants

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

deals or persons

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

 faculté

-0.33

 relative

-0.32

 fontes

-0.32

 penyer

-0.31

Objs

-0.29

 parlant

-0.29

 kematian

-0.29

 gearbeitet

-0.28

шься

-0.28

 plataformas

-0.27

POSITIVE LOGITS

 oprot

0.75

AndEndTag

0.71

AntiForgeryToken

0.65

uxxxx

0.63

addGap

0.61

 ब्रेकडाउन

0.60

 defaultstate

0.60

 الاطلاع

0.59

лтемелер

0.58

webElementXpaths

0.58

Activations Density 0.004%

terms related to job candidates and applicants

deals or persons

No Comments

No Known Activations