INDEX

Explanations

medical and demographic terms

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

non-English words or text elements that refer to people, particularly in Slavic languages.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @neilrathi

This neuron detects the Cyrillic word “пользователь” (“user”), firing strongly on its sub-tokens.

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

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_22/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

to

-1.91

on

-1.84

and

-1.79

by

-1.63

or

-1.62

at

-1.61

in

-1.51

if

-1.46

for

-1.39

 from

-1.27

POSITIVE LOGITS

who

1.62

ᚑ

1.54

 garagem

1.52

 Estructura

1.48

 Consejos

1.45

 Garantía

1.45

 yetu

1.45

 offensichtlich

1.45

 สำหรับ

1.44

 Pantalla

1.43

Activations Density 0.009%

medical and demographic terms

non-English words or text elements that refer to people, particularly in Slavic languages.

This neuron detects the Cyrillic word “пользователь” (“user”), firing strongly on its sub-tokens.

No Comments

No Known Activations

medical and demographic terms

non-English words or text elements that refer to people, particularly in Slavic languages.

This neuron detects the Cyrillic word “пользователь” (“user”), firing strongly on its sub-tokens.

No Comments

No Known Activations