INDEX

Explanations

multilingual text processing

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

demonstrative pronouns in non-English languages (particularly Russian, Japanese, and Chinese), which point to or specify particular things.

oai_token-act-pair · claude-4-5-sonnet Triggered by @bfgcn7frf5

This neuron fires on non-Latin (e.g. Cyrillic, Japanese, Chinese) script tokens.

oai_token-act-pair · o4-mini 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

Not in Any Lists

No Comments

Negative Logits

踅

-2.91

鰱

-2.80

蠓

-2.66

ൎ

-2.63

驯

-2.61

we

-2.53

攆

-2.52

鸲

-2.41



-2.34

觏

-2.23

POSITIVE LOGITS

—

3.56

3.31

也有

2.89

2.78

اااا

2.77

他的

2.67

2.56

大约

2.55

2.48

be

2.42

Activations Density 0.003%

multilingual text processing

demonstrative pronouns in non-English languages (particularly Russian, Japanese, and Chinese), which point to or specify particular things.

This neuron fires on non-Latin (e.g. Cyrillic, Japanese, Chinese) script tokens.

No Comments

No Known Activations

multilingual text processing

demonstrative pronouns in non-English languages (particularly Russian, Japanese, and Chinese), which point to or specify particular things.

This neuron fires on non-Latin (e.g. Cyrillic, Japanese, Chinese) script tokens.

No Comments

No Known Activations