INDEX

Explanations

Objective-C class declarations

The neuron strongly activates on tokens containing digits (i.e. numeric literals or version numbers).

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

Cuánto

-0.81

nikom

-0.80

 zulassen

-0.79

NSUserDefaults

-0.78

pharmacy

-0.75

reactivex

-0.74

IDENTAL

-0.74

Giá

-0.73

 ۱۹

-0.72

вати

-0.71

POSITIVE LOGITS

PF

0.88

ience

0.88

cl

0.85

JS

0.82

ym

0.81

AF

0.80

hư

0.79

MW

0.78

ume

0.77

JK

0.76

Activations Density 0.104%