INDEX

Explanations

Gene | output

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

table column headers and table formatting in text.

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

This neuron detects structural formatting delimiters—especially braces (“{”, “}{”) and pipe characters (“|”) used in arrays or table-like layouts.

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

форма

-0.98

aaaa

-0.98

 дета

-0.93

 winery

-0.92

miy

-0.91

мя

-0.91

lucht

-0.91

 музыки

-0.90

Ậ

-0.90

ของ

-0.89

POSITIVE LOGITS

所说

1.00

 increasing

0.98

过头

0.94

éton

0.94

autorisation

0.92

 Pairs

0.90

Compan

0.90

 häls

0.90

 свої

0.89

 rapidly

0.89

Activations Density 0.003%

Gene | output

table column headers and table formatting in text.

This neuron detects structural formatting delimiters—especially braces (“{”, “}{”) and pipe characters (“|”) used in arrays or table-like layouts.

No Comments

No Known Activations

Gene | output

table column headers and table formatting in text.

This neuron detects structural formatting delimiters—especially braces (“{”, “}{”) and pipe characters (“|”) used in arrays or table-like layouts.

No Comments

No Known Activations