INDEX

Explanations

trained

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

The neuron activates on large numeric tokens—particularly high-magnitude or four-digit numbers (often with decimals).

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

commas separating numerical values or list items in formal or technical writing.

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

unusual typographic and formal markers in text, such as symbols, acronyms/abbreviations, and capitalized proper nouns that stand out in technical or structured contexts.

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

טי

0.82

 successivo

0.78

 bahagian

0.77

 работ

0.77

ção

0.76

 esse

0.76

на

0.75

 acert

0.75

 esses

0.72

 aquele

0.72

POSITIVE LOGITS

 रकम

0.75

CHREIB

0.74

 dehydrogenase

0.73

ಹಾಸ

0.71

 alerts

0.71

氢

0.70

ക്കണം

0.69

 Umesh

0.69

্লেখ

0.68

Hepinize

0.68

Activations Density 0.000%

trained

The neuron activates on large numeric tokens—particularly high-magnitude or four-digit numbers (often with decimals).

commas separating numerical values or list items in formal or technical writing.

unusual typographic and formal markers in text, such as symbols, acronyms/abbreviations, and capitalized proper nouns that stand out in technical or structured contexts.

No Comments

No Known Activations