INDEX

Explanations

features related to programming syntax and structure

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

special characters and programming tokens

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

 ivelany

-0.38

utilisons

-0.30

 مرئيه

-0.29

تقاوى

-0.29

ViewImports

-0.29

 ostavi

-0.28

UnusedPrivate

-0.28

(!__

-0.26

 selben

-0.25

 урна

-0.25

POSITIVE LOGITS

 queſto

0.72

addCriterion

0.66

󠀠

0.62

 ويكيپ

0.62

 パンチラ

0.60

๣

0.60

 linkages

0.59

<unused41>

0.59

 erſten

0.59

[@BOS@]

0.59

Activations Density 0.367%

features related to programming syntax and structure

special characters and programming tokens

No Comments

No Known Activations