INDEX

Explanations

Class<?> reflection

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

-0.11

 Arrow

-0.10

 Registry

-0.10

 Colleg

-0.09

 Hammer

-0.09

mentor

-0.09

 chromat

-0.09

registry

-0.09

isor

-0.09

 èĲ

-0.09

POSITIVE LOGITS

 Class

0.17

sun

0.14

 Reflection

0.13

Dex

0.12

 reflection

0.12

 reflections

0.12

Reflection

0.12

dex

0.11

.newInstance

0.11

DEX

0.11

Activations Density 0.038%