INDEX

Explanations

phrases indicating the existence and presentation of findings or assertions

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

New

-0.34

 Polsek

-0.33

addHeader

-0.32

Bubble

-0.31

 पास

-0.31

The

-0.31

House

-0.30

 ساح

-0.30

settings

-0.30

 Northeast

-0.30

POSITIVE LOGITS

 resourceCulture

0.84

 ChromeDriver

0.72

findpost

0.68

 ModelExpression

0.65

 kasarigan

0.65

PMailer

0.64

 мәкал

0.63

AccessorTable

0.63

WebElementEntity

0.62

 EconPapers

0.60

Activations Density 0.082%

phrases indicating the existence and presentation of findings or assertions

No Comments

No Known Activations