INDEX

Explanations

talking about sometimes

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

 typical

-0.09

 toReturn

-0.08

 )\n\n\n\n\n\n\n\n

-0.08

oka

-0.08

 Hass

-0.08

<br

-0.08

å¿ħé¡»

-0.08

 Disp

-0.08

 :\n\n\n\n

-0.08

POSITIVE LOGITS

 sometimes

0.20

 Sometimes

0.20

Sometimes

0.18

sometimes

0.17

 talking

0.13

 Ð¸Ð½Ð¾Ð³Ð´Ð°

0.12

 Often

0.12

 often

0.12

oft

0.11

 chances

0.11

Activations Density 0.039%