INDEX

Explanations

structured data or formatting elements, particularly related to tables and lists

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

table formatting

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

 FetchType

-0.62

styleType

-0.54

 ddelweddau

-0.52

󠁣

-0.50

OrNil

-0.49

 Italijani

-0.47

webElementXpaths

-0.45

enterOuterAlt

-0.44

 intStringLen

-0.42

 Skyl

-0.42

POSITIVE LOGITS

 defaultstate

0.48

 cherchés

0.48

parametrize

0.46

 &___

0.44

 manzanas

0.43

 Comprometido

0.43

makeConstraints

0.42

 kirke

0.42

 ucapnya

0.41

 fotografico

0.41

Activations Density 0.025%

structured data or formatting elements, particularly related to tables and lists

table formatting

No Comments

No Known Activations