INDEX

Explanations

smoking, smoke, or tobacco use

The main thing this neuron does is detect mentions of smoking or inhaling substances (e.g. tobacco, hookah, vaping, alcohol).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 वस

0.55

เคลื่อน

0.54

 نور

0.52

擠

0.52

 teddy

0.50

teddy

0.50

immobil

0.50

⎞

0.50

щен

0.49

erné

0.49

POSITIVE LOGITS

 tobacco

3.00

 cigarettes

2.89

 cigarette

2.84

 Tobacco

2.84

Tobacco

2.81

 smokers

2.66

 smoking

2.64

 smoker

2.58

 nicotine

2.53

 Smoking

2.53

Activations Density 0.204%