INDEX

Explanations

Â¿En quÃ© puedo ayudarte

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

\n

-0.08

Lan

-0.08

&#8203;

-0.08

Meh

-0.08

ours

-0.08

 fucking

-0.07

 https

-0.07

POSITIVE LOGITS

¶Į

0.12

 addCriterion

0.11

<|begin_of_text|>

0.11

ÂĢÂĢ

0.10

AdapterManager

0.10

įng

0.09

 Erotische

0.09

ARGS

0.09

 -*-č\n

0.09

¦æĥħ

0.09

Activations Density 0.013%