INDEX

Explanations

references to editor-related terms in code or written content

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

editor and code elements

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

่

-0.47

intérêt

-0.47

 positive

-0.47

rawValue

-0.46

 Positive

-0.46

Positive

-0.46

mapStateToProps

-0.45

 Wallace

-0.45

Int

-0.45

 Naik

-0.45

POSITIVE LOGITS

 editor

1.18

 Editor

1.14

editor

1.13

Editor

1.11

 editors

1.00

 EDITOR

1.00

EDITOR

1.00

 UnityEditor

0.94

 Editors

0.92

Editors

0.88

Activations Density 0.005%

references to editor-related terms in code or written content

editor and code elements

No Comments

No Known Activations