INDEX

Explanations

datetime and timestamp

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

 Dread

-0.10

Alv

-0.10

 dipl

-0.10

 Tail

-0.09

 starttime

-0.09

starttime

-0.09

 Graves

-0.09

Alo

-0.09

 clocks

-0.09

 Spring

-0.09

POSITIVE LOGITS

.strftime

0.12

 unaware

0.11

 awareness

0.11

 micro

0.11

.strptime

0.11

aware

0.11

Na

0.10

 naive

0.10

Na

0.10

dt

0.10

Activations Density 0.046%