INDEX

Explanations

legal documents and licenses

The neuron fires on tokens used in opening comment blocks and license‐header boilerplate (e.g. comment delimiters and the “This file is distributed under the terms of the …” text).

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

 both

-0.83

 this

-0.79

激励

-0.79

by

-0.77

 just

-0.77

쿼

-0.77

 while

-0.76

 Spaß

-0.76

 their

-0.75

 también

-0.75

POSITIVE LOGITS

acterium

0.93

צב

0.90

 Mohamad

0.88

准备好

0.87

#+#

0.85

 beantragt

0.84

addressed

0.84

ubo

0.82

 denoting

0.81

TagMode

0.81

Activations Density 0.005%