INDEX

Explanations

instances of the word "new" in programming-related contexts

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

 columnIndex

-0.68

 totalCount

-0.59

 courseId

-0.59

 groupId

-0.57

 statusCode

-0.56

PerformLayout

-0.56

akhstan

-0.56

 admire

-0.56

 loadImage

-0.56

 onFailure

-0.54

POSITIVE LOGITS

new

0.83

New

0.68

NEW

0.65

new

0.60

New

0.60

新

0.57

NEW

0.56

 newOwner

0.55

Nuevo

0.55

 neuen

0.54

Activations Density 0.024%

instances of the word "new" in programming-related contexts

No Comments

No Known Activations