INDEX

Explanations

cigarette smoking

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

terms related to countries, languages, and official organizations or entities.

oai_token-act-pair · claude-3-7-sonnet-20250219

Triggered by @neilrathi

This neuron activates on non-English words (especially German terms).

oai_token-act-pair · o4-mini

Triggered by @jyhe0408

non-English words and scripts, especially salient foreign nouns and vocabulary.

oai_token-act-pair · gpt-5

Triggered by @jyhe0408

words in non-English languages, particularly from Slavic, Germanic, and Asian language texts.

oai_token-act-pair · claude-4-5-sonnet

Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

𐐪

-3.00

ised

-2.81

Ꮏ

-2.66

翃

-2.61

嫆

-2.52

鹦

-2.45

𓐍

-2.45

ed

-2.41

ビック

-2.36

er

-2.33

POSITIVE LOGITS

3.81

2.72

”)

2.55

CHREIB

2.47

 поэтому

2.41

ers

2.39

at

2.33

 verbess

2.31

--;

2.28

ಞ

2.23

Activations Density 0.005%

cigarette smoking

terms related to countries, languages, and official organizations or entities.

This neuron activates on non-English words (especially German terms).

non-English words and scripts, especially salient foreign nouns and vocabulary.

words in non-English languages, particularly from Slavic, Germanic, and Asian language texts.

No Known Activations