INDEX

Explanations

film reviews and dates

The neuron fires on evaluative language about a film’s reception—words and phrases that convey critical or audience judgment (e.g. “mixed reviews,” “positive response,” “struck gold,” “overrated,” etc.).

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

use

-2.38

 during

-1.98

 while

-1.88

 when

-1.88

 used

-1.87

Filmography

-1.82

 useful

-1.82

 όταν

-1.79

 quando

-1.79

 data

-1.77

POSITIVE LOGITS

 herbes

2.20

盉

2.16

 ruines

2.14



2.11

 boya

2.11

銪

2.05

 parfums

1.98

 murs

1.89

 pilotes

1.88

 braço

1.86

Activations Density 0.816%