INDEX

Explanations

mathematical and code delimiters

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

The neuron is highly responsive to numeric tokens—especially floating‐point or citation‐style numbers—flagging occurrences of numeric data in the text.

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

closing parentheses followed by digits in academic or technical citations.

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

mentions of decimal numbers (floating-point numeric values), often appearing in citations or technical text.

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

emple

0.82

ко

0.82

ate

0.80

eng

0.80

sync

0.79

gu

0.78

az

0.78

dest

0.78

square

0.76

gra

0.74

POSITIVE LOGITS

 Sungai

0.93

 ໜອງ

0.91

 大人

0.88

ጷ

0.87

断

0.81

 Fian

0.80

 windfall

0.80

 overseen

0.80

 thre

0.80

 Возможно

0.79

Activations Density 0.000%

mathematical and code delimiters

The neuron is highly responsive to numeric tokens—especially floating‐point or citation‐style numbers—flagging occurrences of numeric data in the text.

closing parentheses followed by digits in academic or technical citations.

mentions of decimal numbers (floating-point numeric values), often appearing in citations or technical text.

No Comments

No Known Activations