INDEX

Explanations

code comments and licenses

The neuron responds to alphanumeric identifier tokens (e.g. mixed-letter/digit names like “ItemID”).

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

 był

-2.33

𖥦

-2.23

")]

-2.23

breviations

-2.22

 était

-2.22

 был

-2.14

er

-2.11

</h2>

-2.08

ቍ

-2.02

which

-2.00

POSITIVE LOGITS

</i>

3.08

</h1>

2.55

2.28

蹰

2.27

</sub>

2.23

 otros

2.19

 strana

2.16

籥

2.08

 outros

2.05

跺

2.00

Activations Density 0.000%