INDEX

Explanations

time-consuming

The neuron detects mentions of tasks or processes described as “time-consuming.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

能在

-2.17

Despite

-2.00

Aside

-1.90

Numerous

-1.87

Because

-1.86

 trovo

-1.82

Such

-1.82

Since

-1.80

Besides

-1.77

For

-1.74

POSITIVE LOGITS

鸶

2.45

玙

2.31

珰

2.23

 trám

2.17

汭

2.13

 obrazy

2.08



2.08

 niej

2.03

ize

2.03

塍

2.02

Activations Density 0.009%