INDEX

Explanations

verbs and actions related to processes, training, and modifications

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-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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

:✨

-0.76

contentLoaded

-0.68

ruptedException

-0.65

migrationBuilder

-0.61

 AppModule

-0.58

enumii

-0.56

EndGlobalSection

-0.54

ioutil

-0.54

__((

-0.53

usercontent

-0.53

POSITIVE LOGITS

vœ

0.46

 themselves

0.39

 unify

0.38

ly

0.37

 nueces

0.36

 maksi

0.36

 Moskau

0.36

 désert

0.36

 borra

0.36

วิว

0.35

Activations Density 0.085%

verbs and actions related to processes, training, and modifications

No Comments

No Known Activations