INDEX

Explanations

special characters or formatting used in text

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Madura

-0.40

 opportunity

-0.39

 quæ

-0.36

 cuad

-0.35

 Doğ

-0.35

 ejus

-0.35

 Esau

-0.34

Club

-0.34

KID

-0.34

 Clubs

-0.34

POSITIVE LOGITS

1.71

AnimationsModule

0.57

 oprot

0.56

 récentes

0.51

 Schrö

0.51

AnchorStyles

0.51

PGP

0.50

poptotic

0.50

 récente

0.50

 roulant

0.50

Activations Density 0.000%

special characters or formatting used in text

No Comments

No Known Activations