INDEX

Explanations

possibly

The neuron strongly activates on the hedge adverb “possibly” (in any casing).

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

嚨

-2.14

箆

-2.14

 TÉCNICA

-2.02

四处

-1.94

tedrale

-1.94

)”

-1.94

😵

-1.94

uminazione

-1.92

🍩

-1.90

">{{

-1.88

POSITIVE LOGITS

峮

2.56

,\\

2.44

 場所

2.39

 termed

2.34

 regarded

2.34

 trám

2.27

 一个

2.09

횃

2.09

珰

2.06



2.06

Activations Density 0.003%