INDEX

Explanations

dates and timestamps

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

date stamps

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

 Wachs

-0.41

 Kowal

-0.37

 Aufenthalts

-0.35

 semula

-0.34

 utilizarse

-0.34

Personendaten

-0.34

 tarjetas

-0.33

 Sahne

-0.33

 Kleidung

-0.33

 adentro

-0.33

POSITIVE LOGITS

 CanadaChoose

0.66

enumii

0.62

Personensuche

0.60

twimg

0.57

enterOuterAlt

0.57

RegressionTest

0.56

참고

0.54

Portail

0.52

 estekak

0.51

Kjelder

0.50

Activations Density 0.025%

dates and timestamps

date stamps

No Comments

No Known Activations

dates and timestamps

date stamps

No Comments

No Known Activations