INDEX

Explanations

assistant role identifier

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

wow

-0.10

ober

-0.09

 rhetoric

-0.09

Wow

-0.09

arness

-0.09

IGH

-0.09

bon

-0.09

 roller

-0.09

onn

-0.08

atro

-0.08

POSITIVE LOGITS

 screw

0.18

 Screw

0.16

Bah

0.15

bah

0.14

who

0.13

Bah

0.13

 cares

0.12

bah

0.12

shr

0.11

who

0.11

Activations Density 0.051%