INDEX

Explanations

code commands and foreign words

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

line breaks or whitespace that separate distinct sections or paragraphs in structured text.

oai_token-act-pair · claude-4-5-sonnet Triggered by @bfgcn7frf5

The neuron flags document section headings or metadata labels (e.g. “Abstract:”, “About this item”, “Q:”), i.e. lines that mark or introduce structured sections.

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

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/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

 often

-1.66

 only

-1.61

our

-1.48

 That

-1.41

 любит

-1.33

 Often

-1.33

 touted

-1.33

so

-1.30

 where

-1.29

 pajero

-1.24

POSITIVE LOGITS

↵

1.95

訁

1.80

 wspania

1.43

 ďal

1.36

 mają

1.36

ومن

1.34

’

1.33

ৄ

1.33

hôte

1.32

sandalia

1.31

Activations Density 0.001%

code commands and foreign words

line breaks or whitespace that separate distinct sections or paragraphs in structured text.

The neuron flags document section headings or metadata labels (e.g. “Abstract:”, “About this item”, “Q:”), i.e. lines that mark or introduce structured sections.

No Comments

No Known Activations

code commands and foreign words

line breaks or whitespace that separate distinct sections or paragraphs in structured text.

The neuron flags document section headings or metadata labels (e.g. “Abstract:”, “About this item”, “Q:”), i.e. lines that mark or introduce structured sections.

No Comments

No Known Activations