INDEX

Explanations

enough or sufficient

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

 excessively

-0.10

 adequately

-0.09

keh

-0.09

 deter

-0.08

akin

-0.08

 moderately

-0.08

ofi

-0.08

 identical

-0.08

 thoroughly

-0.08

eldon

-0.08

POSITIVE LOGITS

 enough

1.19

 Enough

0.88

 Äĳá»§

0.66

 sufficient

0.62

Enough

0.57

è¶³

0.56

å¤Ł

0.52

ufficient

0.51

 Ð´Ð¾ÑģÑĤÐ°ÑĤ

0.48

 dostate

0.45

Activations Density 0.246%