INDEX

Explanations

names and identifiers related to programming and user information

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

Indian and European male names

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

ViewFeatures

-0.71

 AssemblyCulture

-0.66

 kaarangay

-0.65

OGND

-0.65

ckså

-0.64

 Italijani

-0.60

ThroughAttribute

-0.57

Autoritní

-0.57

tvguidetime

-0.57

writeFieldEnd

-0.56

POSITIVE LOGITS

 الحره

0.37

errHandler

0.35

ignite

0.33

Ney

0.31

MultipartFile

0.31

Mol

0.30

mis

0.29

الإنجليزية

0.29

 Miche

0.29

Bri

0.29

Activations Density 0.006%

names and identifiers related to programming and user information

Indian and European male names

No Comments

No Known Activations