INDEX

Explanations

cache entry, factory, manager, ttl, collections

The neuron specifically detects mentions of caching (tokens containing “cache” or “cached”) in 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

婓

-2.17

』、

-2.16

 folosi

-1.82

媄

-1.80

ра

-1.73

sthe

-1.71

泐

-1.69

淠

-1.69

indsay

-1.66

 dévelo

-1.63

POSITIVE LOGITS

<bos>

2.00

in

1.87

 комиссии

1.67

𞤢

1.64

ᯅ

1.62

 hinder

1.60

 phrase

1.58

1.57

 дорогой

1.53

 να

1.51

Activations Density 0.014%