INDEX

Explanations

less than 0

This neuron responds to tokens in mathematical expressions—especially Greek‐letter variables (like ΔG) and inequality comparisons.

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

)<\

-0.98

}<\

-0.82

<\

-0.76

Portály

-0.73

 learn

-0.72

 ظاهر

-0.71

 anhydrous

-0.71

owsky

-0.70

 Heft

-0.69

 たま

-0.69

POSITIVE LOGITS

 below

2.06

Below

1.59

below

1.53

 BELOW

1.48

 Below

1.47

negative

1.32

 ниже

1.30

 negative

1.29

 abaixo

1.19

()<

1.13

Activations Density 0.015%