INDEX

Explanations

sentiment and mentions

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

 statistics

-0.11

je

-0.11

Statistics

-0.10

 Grove

-0.10

HM

-0.09

 statistical

-0.09

 diagn

-0.09

 Courage

-0.09

 Statistics

-0.09

azel

-0.09

POSITIVE LOGITS

 sentiment

0.25

 Sent

0.20

 sentiments

0.18

Sent

0.18

sent

0.16

 sentimental

0.16

 mention

0.15

èĪ

0.14

 mentions

0.14

 text

0.14

Activations Density 0.195%