INDEX

Explanations

conjunctions and clauses

The neuron chiefly activates on numeric tokens or expressions of numerical values (e.g. measurements, years, quantities).

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

atzes

-1.34

riezmann

-1.29

</b>

-1.23

⬮

-1.21

 Mongols

-1.19

edf

-1.18

Raptor

-1.16

Rango

-1.16

Alike

-1.15

 выясни

-1.11

POSITIVE LOGITS

now

1.52

times

1.37

ном

1.35

،

1.34

een

1.29

 charmant

1.29

 then

1.27

 teraz

1.27

 have

1.23

 aperto

1.23

Activations Density 0.102%