INDEX

Explanations

mathematical expressions and expressions that indicate relationships or operations between elements

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

mathematical expressions and latex

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

Dichloropropane

-0.37

 الرياضيه

-0.37

primera

-0.33

avajillas

-0.33

 Sock

-0.32

 Dryden

-0.32

 Fußballspieler

-0.32

RESUMO

-0.31

 dunque

-0.31

 ennen

-0.31

POSITIVE LOGITS

transQ

0.63

帖最后由

0.59

 PeEnEo

0.59

URLException

0.56

XtraReports

0.56

EndProject

0.53

 Taktlose

0.52

تقاوى

0.52

 estekak

0.51

enumi

0.50

Activations Density 0.118%

mathematical expressions and expressions that indicate relationships or operations between elements

mathematical expressions and latex

No Comments

No Known Activations