INDEX

Explanations

proper nouns, specifically names of authors and contributors

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

names and social/political concepts

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

 autorytatywna

-0.51

 surla

-0.50

 kaarangay

-0.50

 للاسماء

-0.46

webElementXpaths

-0.46

LookAnd

-0.45

verifyException

-0.45

GEBURTSDATUM

-0.44

-------------</

-0.44

 povr

-0.43

POSITIVE LOGITS

 sociologist

0.56

 Social

0.52

 social

0.52

 socio

0.49

Social

0.48

 socialist

0.46

 Garnet

0.45

Gus

0.45

 participatory

0.45

 political

0.44

Activations Density 0.037%

proper nouns, specifically names of authors and contributors

names and social/political concepts

No Comments

No Known Activations