INDEX

Explanations

Ira, Nia, TLR, Bauhaus, VEGF

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

first names or personal identifiers in text, particularly those that appear in formal contexts like legal documents, publications, or technical writing.

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

The neuron fires on isolated capitalized tokens—proper nouns, personal names, or technical class/package identifiers.

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

鞮

-2.98

᭝

-2.97

งิน

-2.95

 helicópter

-2.95

雠

-2.92

продам

-2.89

膑

-2.86

 Comprometido

-2.86

ſam

-2.81

 miniaturka

-2.78

POSITIVE LOGITS

3.47

3.44

“

3.33

is

3.19

in

3.09

It

2.97

as

2.91

2.88

но

2.86

by

2.83

Activations Density 0.017%

Ira, Nia, TLR, Bauhaus, VEGF

first names or personal identifiers in text, particularly those that appear in formal contexts like legal documents, publications, or technical writing.

The neuron fires on isolated capitalized tokens—proper nouns, personal names, or technical class/package identifiers.

No Comments

No Known Activations

Ira, Nia, TLR, Bauhaus, VEGF

first names or personal identifiers in text, particularly those that appear in formal contexts like legal documents, publications, or technical writing.

The neuron fires on isolated capitalized tokens—proper nouns, personal names, or technical class/package identifiers.

No Comments

No Known Activations