INDEX

Explanations

specie or released

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

HTML-like formatting tags and markup elements in text.

oai_token-act-pair · claude-3-7-sonnet-20250219

Triggered by @neilrathi

The neuron fires on markup or formatting tokens (e.g. asterisks, underscores, HTML/Markdown tags) regardless of the surrounding content.

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

Negative Logits

</strong>

-2.45

⿷

-1.95

</h4>

-1.86

呖

-1.82

-1.71

ぢ

-1.70

ませて

-1.62

 ſtre

-1.61

 mépris

-1.50

㩒

-1.50

POSITIVE LOGITS

is

2.09

ༀ

1.60

可惜

1.53

有一

1.50

鮭

1.49

ꦒ

1.48

This

1.46

 there

1.46

Jereo

1.45

how

1.44

Activations Density 0.001%

specie or released

HTML-like formatting tags and markup elements in text.

The neuron fires on markup or formatting tokens (e.g. asterisks, underscores, HTML/Markdown tags) regardless of the surrounding content.

No Known Activations

specie or released

HTML-like formatting tags and markup elements in text.

The neuron fires on markup or formatting tokens (e.g. asterisks, underscores, HTML/Markdown tags) regardless of the surrounding content.

No Known Activations