INDEX

Explanations

actions following conjunctions or pronouns

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

akk

-0.10

Intialized

-0.09

æĭĶ

-0.08

cribe

-0.08

|:

-0.08

adeon

-0.08

fromJson

-0.08

èªĮ

-0.08

interop

-0.08

 therap

-0.08

POSITIVE LOGITS

 start

0.10

 philosoph

0.10

 went

0.10

pam

0.09

 memor

0.09

 barg

0.09

 Became

0.09

 dissert

0.08

 became

0.08

Activations Density 0.191%