INDEX

Explanations

syntax elements related to programming and code structure

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

mathematical symbols

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

bootstrapcdn

-0.61

mapStateToProps

-0.59

 onOptions

-0.51



-0.50

parsedMessage

-0.47

oredCriteria

-0.46

mergeFrom

-0.46

 protoimpl

-0.45

WaitGroup

-0.45

elemField

-0.44

POSITIVE LOGITS

 թվական

0.42

 usher

0.42

Bri

0.41

لينكات

0.41

MTV

0.39

pagestyle

0.39

 unſ

0.39

brtc

0.38

ISTS

0.38

Sha

0.38

Activations Density 0.032%

syntax elements related to programming and code structure

mathematical symbols

No Comments

No Known Activations