INDEX

Explanations

not responding

The neuron strongly activates on content-bearing, information‐rich words (like nouns and verbs) and not on common function words.

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

 querer

-0.99

 battered

-0.96

 usar

-0.95

 queda

-0.93

 vouloir

-0.92

 detrimental

-0.91

 will

-0.91

 struggle

-0.90

 planned

-0.90

use

-0.89

POSITIVE LOGITS

正在

1.55

 đang

1.50

กำลัง

1.17

 doing

1.13

 performing

1.11

currently

1.11

 enjoying

1.09

 happening

1.05

 currently

1.05

 running

1.04

Activations Density 0.104%