INDEX

Explanations

terms related to jail procedures and inmate grievances

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

inmate telephones

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

vermögen

-0.39

 felicidade

-0.38

icksburg

-0.35

anzia

-0.32

 Evan

-0.32

PROPERTIES

-0.32

 campagna

-0.31

uttavia

-0.31

 bangsa

-0.31

cellona

-0.30

POSITIVE LOGITS

 inmates

0.72

 inmate

0.68

 prisoners

0.68

 Prisoners

0.63

 prisoner

0.60

prison

0.59

 ſche

0.57

 ***!

0.57

 prison

0.55

 autorytatywna

0.53

Activations Density 0.118%

terms related to jail procedures and inmate grievances

inmate telephones

No Comments

No Known Activations