INDEX

Explanations

predicting "when"

introduces when clauses

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

িন্দ

0.37

ത്തര

0.35

úrate

0.35

ତି

0.34

 আপাতত

0.34

ಟ್‌

0.33

ณฑ

0.32

 နှစ်

0.32

ਵੀ

0.32

λι

0.31

POSITIVE LOGITS

 when

4.75

when

4.59

 cuando

4.41

 когда

4.41

 quando

4.34

 när

4.06

 عندما

4.03

 когато

4.00

 όταν

3.95

 WHEN

3.94

Activations Density 2.065%