INDEX

Explanations

The neuron fires on numeric literals (especially multi‐digit or decimal numbers).

It detects large floating-point numeric tokens (decimal numbers like 1600–1680 with several digits).

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 during

0.37

and

0.37

 parcels

0.33

at

0.32

 மற்றும்

0.32

 from

0.32

 occasionally

0.31

 regularly

0.31

 locally

0.30

 workers

0.30

POSITIVE LOGITS

 esimerk

0.38

 problemática

0.37

 teorías

0.36

 ಸಮಸ್ಯ

0.36

 refute

0.36

 этого

0.35

 explicación

0.35

 conundrum

0.35

 ponencia

0.35

 overfitting

0.35

Activations Density 0.000%