INDEX

Explanations

references to interviews and media outlets

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

media outlets conducting interviews

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

 ब्रेकडाउन

-0.60

 AssemblyCulture

-0.58

 незавершена

-0.56

KommentareTeilen

-0.55

 disambiguazione

-0.54

ollectionView

-0.54

 PagesJaunes

-0.54

 auroit

-0.53

IntoConstraints

-0.53

 avoient

-0.52

POSITIVE LOGITS

 interview

0.54

Interview

0.52

interview

0.50

 Interview

0.50

 interviewer

0.47

simum

0.45

 Interviews

0.42

Interviews

0.41

litter

0.41

 interviews

0.41

Activations Density 0.009%

references to interviews and media outlets

media outlets conducting interviews

No Comments

No Known Activations