INDEX

Explanations

evaluation of outcomes

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ä¸¥

-0.09

/inc

-0.09

Ð´Ð°Ð»

-0.09

éļĥ

-0.09

åīįæīĢæľª

-0.08

 Ð·Ð°ÑīÐ¸ÑĤ

-0.08

Ø¹Ø±Ø¨

-0.08

Rig

-0.08

ock

-0.08

åĲĳå¾Ģ

-0.08

POSITIVE LOGITS

è°ĥåĳ³

0.11

çģµæ°Ķ

0.10

ExecutionContext

0.10

ä¸Ģéĥ¨

0.10

çªģåĩºéĹ®é¢ĺ

0.10

çļĦæĦŁæĥħ

0.10

æĸ¯é¡¿

0.10

æĦŁäºº

0.10

æĦŁæĥħ

0.10

çĬ

0.09

Activations Density 0.065%