INDEX

Explanations

new connections, places, money

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

(íģ¬ê¸°

-0.09

 embargo

-0.09

imore

-0.09

-0.08

<|begin_of_text|>

-0.08

ingly

-0.08

.Formatter

-0.08

¨ë¶Ģ

-0.08

çĲĨçĶ±

-0.08

anio

-0.08

POSITIVE LOGITS

 behavior

0.09

/new

0.08

 ideas

0.08

anki

0.08

 outcome

0.08

anders

0.08

 Cham

0.08

heid

0.08

 information

0.08

odies

0.08

Activations Density 0.202%