INDEX

Explanations

numbers followed by 's'

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

This neuron is especially sensitive to segments of first-person personal narratives or anecdotes.

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

plural people groups**

np_max-act-logits · claude-4-5-sonnet Triggered by @jrosseruk

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

 digestible

0.79

 catalogs

0.74

а

0.73

 gắn

0.71

 intersects

0.69

 dispara

0.69

 uploads

0.69

这类

0.69

 kullanılır

0.69

 proliferate

0.68

POSITIVE LOGITS

全員

1.31

 جميعا

1.24

 themselves

1.20

各自

1.20

それぞれの

1.11

 كلهم

1.11

それぞれ

1.04

三人

1.03

 それぞれ

1.02

 estaban

1.00

Activations Density 0.783%

numbers followed by 's'

This neuron is especially sensitive to segments of first-person personal narratives or anecdotes.

plural people groups**

No Comments

No Known Activations

numbers followed by 's'

This neuron is especially sensitive to segments of first-person personal narratives or anecdotes.

plural people groups**

No Comments

No Known Activations