INDEX

Explanations

references to specific television and radio stations

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

radio and tv station call signs

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

UnusedPrivate

-0.81

LookAnd

-0.76

XmlAccessorType

-0.63

 Photocase

-0.63

 مرئيه

-0.62

makeConstraints

-0.61

 мәкал

-0.60

tagHelperRunner

-0.60

centiles

-0.58

 resourceCulture

-0.57

POSITIVE LOGITS

JRadioButton

0.36

DataOutputStream

0.30

 setDescription

0.30

mbggenerated

0.29

JButton

0.28

 Channel

0.27

 miejsco

0.27

UIButton

0.27

 nearby

0.26

 sumpay

0.26

Activations Density 0.007%

references to specific television and radio stations

radio and tv station call signs

No Comments

No Known Activations