INDEX

Explanations

describing chemical properties and reactions

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

 condem

-0.09

¶Į

-0.08

:\/\/

-0.08

 representatives

-0.08

 legis

-0.08

éħ¸

-0.08

 rational

-0.08

 Scheme

-0.08

 Representatives

-0.08

 Metals

-0.08

POSITIVE LOGITS

 reaction

0.15

 chemical

0.15

 process

0.14

åıįåºĶ

0.13

 react

0.11

 processes

0.11

reaction

0.11

 reactions

0.11

Raw

0.11

 molecular

0.11

Activations Density 0.130%