INDEX

Explanations

transition between states

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

Negative Logits

 deterministic

-0.13

 Stap

-0.10

VIP

-0.09

(||

-0.09

 sculpt

-0.09

ether

-0.09

 stochastic

-0.09

æĳĨ

-0.08

å¥ĩ

-0.08

 Rogers

-0.08

POSITIVE LOGITS

 Mark

0.27

Mark

0.21

 transition

0.20

 Transition

0.19

 transitions

0.19

 state

0.18

transition

0.18

Transition

0.17

 Hidden

0.17

.transition

0.16

Activations Density 0.066%

No Known Activations