INDEX

Explanations

from

The neuron activates on numeric tokens—especially measurements or quantities (including decimals) embedded 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

-2.06

-1.45

-1.40

mathrm

-1.36

fy

-1.29

[]:

-1.28

-1.27

你

-1.23

('

-1.22

<code>

-1.21

POSITIVE LOGITS

 Эти

1.55

〮

1.49

 solusi

1.49

Slayer

1.44

 takich

1.44

 beträ

1.42

 deoarece

1.42

dali

1.42

玙

1.41

 Những

1.40

Activations Density 0.089%