INDEX

Explanations

if followed by condition

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

 setattr

-0.09

(SQLException

-0.09

 Denn

-0.09

iral

-0.09

 retros

-0.08

 gener

-0.08

 frontal

-0.08

ipment

-0.08

FromBody

-0.08

ullo

-0.08

POSITIVE LOGITS

aces

0.13

ToPoint

0.12

rames

0.11

 ëĭ¤ìļ´ë°Ľê¸°

0.11

fy

0.10

((

0.10

(!

0.10

_,

0.10

Activations Density 0.014%