INDEX

Explanations

result

This neuron fires on descriptive terms that denote material or product properties—especially strength, durability, coherence, thickness, or other quality/performance attributes.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 kapat

0.52

̷

0.51

isance

0.50

allax

0.49

ድረግ

0.48

irting

0.48

spruch

0.48

 kapan

0.47

 weiteres

0.47

 активности

0.46

POSITIVE LOGITS

 resulting

3.19

 resultant

3.00

resulting

2.89

 result

2.80

 resultante

2.80

结果

2.69

結果

2.66

 resulted

2.58

 resultado

2.55

result

2.53

Activations Density 0.814%