INDEX

Explanations

references to programming packages

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 canst

-0.45

twimg

-0.41

WebElementEntity

-0.40

 herinner

-0.36

 juiz

-0.34

AndEndTag

-0.34

AsUp

-0.34

נוּ

-0.34

 höchsten

-0.33

yyj

-0.32

POSITIVE LOGITS

package

0.95

 package

0.90

 Package

0.90

Package

0.83

 Packages

0.78

packageName

0.76

 packageName

0.76

 PACKAGE

0.76

 packages

0.73

PackageName

0.71

Activations Density 0.015%

references to programming packages

No Comments

No Known Activations