INDEX

Explanations

terms related to secondhand smoke and exposure risks

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

vaping

np_acts-logits-general · gemini-2.0-flash

vape or vapor

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

reception

-0.62

 réception

-0.59

 出版

-0.59

 لينك

-0.58

 تضيفلها

-0.58

INSEE

-0.57

 مرئيه

-0.57

 HasFactory

-0.57

 Gypsum

-0.57

 Reception

-0.56

POSITIVE LOGITS

 vapor

0.94

 vape

0.93

 vaping

0.88

 vapour

0.81

 Vapor

0.79

 Vape

0.78

vapor

0.78

Vapor

0.76

Vap

0.74

 atom

0.74

Activations Density 0.062%

terms related to secondhand smoke and exposure risks

vaping

vape or vapor

No Comments

No Known Activations