INDEX

Explanations

team contexts

This neuron detects occurrences of the word “team” (especially in identifiers like team_id or class names) in code or structured data.

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

but

-2.86

Of

-2.31

We

-2.25

."

-2.20

 along

-2.05

no

-1.99

")

-1.95

as

-1.95

と言

-1.90

ги

-1.88

POSITIVE LOGITS

 seksi

2.42

Because

2.41

 rajut

2.33

卝

2.25

These

2.20

 gantung

2.19

 artistico

2.16

Ꮉ

2.11

 laku

2.11

 اشار

2.11

Activations Density 0.004%