INDEX

Explanations

future outcomes and planning

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 âĸ³

-0.10

 latest

-0.10

unger

-0.10

éħ

-0.09

otta

-0.09

current

-0.09

íĭ

-0.09

GEST

-0.09

.TestTools

-0.09

 current

-0.09

POSITIVE LOGITS

-proof

0.21

 generations

0.21

proof

0.20

/current

0.18

 tense

0.15

 directions

0.12

 editions

0.12

Proof

0.12

/up

0.12

 proof

0.12

Activations Density 0.037%