INDEX

Explanations

excessive swearing

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

The neuron fires on numeric tokens (especially multi‐digit numbers and decimals), i.e. values and measurements.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

line breaks or paragraph boundaries in business/commercial text.

oai_token-act-pair · claude-4-5-sonnet Triggered by @jyhe0408

expressions that emphasize a wide variety or assortment of things.

oai_token-act-pair · gpt-5 Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tex

0.78

alt

0.77

cido

0.76

al

0.75

github

0.72

Work

0.71

cor

0.71

cu

0.70

↵

0.70

cir

0.70

POSITIVE LOGITS

あるいは

0.88

 каждой

0.86

ным

0.84

 каждая

0.83

ﺄ

0.83

 вы

0.82

 этой

0.80

лением

0.80

 каждому

0.80

 каждый

0.79

Activations Density 0.001%

excessive swearing

The neuron fires on numeric tokens (especially multi‐digit numbers and decimals), i.e. values and measurements.

line breaks or paragraph boundaries in business/commercial text.

expressions that emphasize a wide variety or assortment of things.

No Comments

No Known Activations

excessive swearing

The neuron fires on numeric tokens (especially multi‐digit numbers and decimals), i.e. values and measurements.

line breaks or paragraph boundaries in business/commercial text.

expressions that emphasize a wide variety or assortment of things.

No Comments

No Known Activations