INDEX

Explanations

scores and outcomes

This neuron detects numeric tokens—scores, statistics, and other multi‐digit or decimal number sequences in sports summaries.

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

وردی

-1.42

who

-1.38

鲻

-1.35

男主

-1.29

Nội

-1.27

มาะ

-1.26

 tumult

-1.24

 bewilder

-1.24

 ayudó

-1.24

当你

-1.23

POSITIVE LOGITS

1.54

1.40

 against

1.35

”

1.35

ционер

1.34

was

1.33

de

1.33

=“

1.32

 overall

1.30

Activations Density 0.030%