INDEX

Explanations

JavaScript or AngularJS code structures and syntax

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

code and user interactions

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

 InputDecoration

-0.81

 EconPapers

-0.73

 BoxDecoration

-0.72

 disambiguazione

-0.71

AccessorTable

-0.66

MessageOf

-0.65

AsUp

-0.65

featureID

-0.62

 ExecuteAsync

-0.59

#+#

-0.57

POSITIVE LOGITS

dikan

0.35

 technique

0.34

 personnel

0.33

 logistics

0.33

 techniques

0.33

 asmen

0.33

みなさん

0.32

SneakyThrows

0.31

enschaft

0.31

皆さん

0.31

Activations Density 0.005%

JavaScript or AngularJS code structures and syntax

code and user interactions

No Comments

No Known Activations