INDEX

Explanations

list items followed by comma or newline

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

rego

-0.10

ľĺ

-0.09

ons

-0.08

 Aster

-0.08

Neh

-0.08

 remnants

-0.08

iasi

-0.08

encion

-0.08

jsc

-0.08

aja

-0.08

POSITIVE LOGITS

abra

0.09

 Navigator

0.08

Viv

0.08

éĽ¶

0.08

 Lester

0.08

))==

0.08

ãĢģãĢģ

0.08

tal

0.08

 liner

0.08

viv

0.08

Activations Density 0.134%