INDEX

Explanations

closing code blocks and statements

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

reg

-0.10

udo

-0.09

ing

-0.09

acre

-0.09

agr

-0.09

âĤ

-0.08

rue

-0.08

heit

-0.08

 kits

-0.08

POSITIVE LOGITS

è¼

0.11

VisualStyle

0.10

GC

0.10

.MixedReality

0.10

.updateDynamic

0.10

-------------</

0.10

_vlog

0.10

.Undef

0.10

 endwhile

0.10

>NN

0.10

Activations Density 0.009%