INDEX

Explanations

acronyms and abbreviations related to scientific terminology or processes

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

Personendaten

-0.60

 kasarigan

-0.52

Билгалдахарш

-0.52

 EconPapers

-0.52



-0.51

readFileSync

-0.46

WebElementEntity

-0.44

runApp

-0.43

ppuden

-0.43

 الرياضيه

-0.42

POSITIVE LOGITS

Revenir

0.45

tagHelperRunner

0.45

colgante

0.43

 ModelExpression

0.43

 ankaŭ

0.43

rungsseite

0.42

 noemen

0.40

なって

0.40

 moeite

0.40

 MainAxisSize

0.39

Activations Density 0.110%

acronyms and abbreviations related to scientific terminology or processes

No Comments

No Known Activations