INDEX

Explanations

2000 words in chemical industry

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

_shuffle

-0.10

 sanction

-0.09

amedi

-0.09

ë¹

-0.09

 ØµØ§ØŃ

-0.08

 mbox

-0.08

 Plot

-0.08

á»©t

-0.08

ngo

-0.08

aments

-0.08

POSITIVE LOGITS

 mesh

0.14

 Mesh

0.12

CAS

0.12

mesh

0.11

mg

0.10

 Affairs

0.10

 purity

0.10

loh

0.09

Ih

0.09

 rall

0.09

Activations Density 0.067%