INDEX

Explanations

Times, sessions, MM/DD, dot

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

This neuron lights up on key nouns that serve as section or list headings—terms like “days,” “portions,” “frequency,” “paragraph,” “period,” or “materials” that typically introduce discrete items or steps.

oai_token-act-pair · o4-mini

Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

Afd

0.68

 lama

0.59

ኋላ

0.58

 fxaa

0.57

 unang

0.54

}\,\

0.53

 urination

0.52

 bonjour

0.52

 lidar

0.52

ULT

0.51

POSITIVE LOGITS

庐

0.58

settes

0.56

廬

0.55

cion

0.54

astic

0.54

ುತ್ತದೆ

0.54

ariales

0.54

ambahan

0.54

mal

0.53

 अहम

0.53

Activations Density 0.000%

Times, sessions, MM/DD, dot

This neuron lights up on key nouns that serve as section or list headings—terms like “days,” “portions,” “frequency,” “paragraph,” “period,” or “materials” that typically introduce discrete items or steps.

No Known Activations

Times, sessions, MM/DD, dot

This neuron lights up on key nouns that serve as section or list headings—terms like “days,” “portions,” “frequency,” “paragraph,” “period,” or “materials” that typically introduce discrete items or steps.

No Known Activations

**Times**, **sessions**, **MM**/DD, **dot**

This neuron lights up on key nouns that serve as section or list headings—terms like “days,” “portions,” “frequency,” “paragraph,” “period,” or “materials” that typically introduce discrete items or steps.

No Known Activations

**Times**, **sessions**, **MM**/DD, **dot**

This neuron lights up on key nouns that serve as section or list headings—terms like “days,” “portions,” “frequency,” “paragraph,” “period,” or “materials” that typically introduce discrete items or steps.

No Known Activations

Times, sessions, MM/DD, dot

Times, sessions, MM/DD, dot