INDEX

Explanations

rules for decision making

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

 Hann

-0.09

éĻĲåĪ¶

-0.09

 Monitoring

-0.09

Statics

-0.09

.RunWith

-0.09

andler

-0.08

 segmented

-0.08

_restrict

-0.08

ivy

-0.08

 Wass

-0.08

POSITIVE LOGITS

 decision

0.21

 rule

0.19

 determination

0.17

 rules

0.16

decision

0.16

 criteria

0.16

 Decision

0.16

 Rule

0.15

 determining

0.15

 decisions

0.14

Activations Density 0.176%