INDEX

Explanations

inquiries about grammar

The neuron fires on explicit references to grammar and linguistic structure—e.g. terms like “grammar,” “structure,” “sentences,” or “underlying” in a meta-linguistic discussion.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 получите

-0.82

MatButtonModule

-0.78

 letra

-0.76

textured

-0.76

 Lyrics

-0.76

ætte

-0.75

 Klavier

-0.73

NCR

-0.71

Ț

-0.70

)(((

-0.69

POSITIVE LOGITS

 grammar

2.77

 grammatical

2.52

 gram

2.30

 Grammar

2.27

grammar

2.17

 gramm

2.02

Grammar

1.98

gram

1.89

 Gram

1.88

 грам

1.84

Activations Density 0.062%