INDEX

Explanations

coding keywords and structure in a programming context

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

public class/type declarations

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

Personendaten

-0.85

parsedMessage

-0.78

exitRule

-0.68

ChildScrollView

-0.65

énario

-0.65

 itſelf

-0.62



-0.62

tagext

-0.62

 ModelExpression

-0.59

 becauſe

-0.58

POSITIVE LOGITS

0.54

0.53

the

0.49

an

0.47

 black

0.46

fan

0.46

msub

0.46

 twin

0.45

 transparent

0.44

 French

0.44

Activations Density 0.000%

coding keywords and structure in a programming context

public class/type declarations

No Comments

No Known Activations