INDEX

Explanations

code-related keywords, particularly those associated with module exports and function declarations

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

JavaScript code declaration

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

 eben

-0.38

enumi

-0.36

UrlEncoded

-0.36

ContentAlignment

-0.34

 seduta

-0.33

 сообща

-0.33

findpost

-0.32

wpi

-0.32

ereich

-0.32

 Espinoza

-0.32

POSITIVE LOGITS

AndEndTag

0.87

rungsseite

0.63

 AssemblyTitle

0.62

ValueStyle

0.58

 &___

0.57

AddTagHelper

0.57

 CreateTagHelper

0.54

BASEPATH

0.52

OGND

0.52

KommentareTeilen

0.52

Activations Density 0.004%

code-related keywords, particularly those associated with module exports and function declarations

JavaScript code declaration

No Comments

No Known Activations