INDEX

Explanations

subtracting and isolating

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

 Altern

-0.11

 âĪļ

-0.10

 Aber

-0.09

âĪļ

-0.09

imore

-0.09

 Ster

-0.09

 spell

-0.09

tern

-0.09

 tang

-0.09

hatt

-0.09

POSITIVE LOGITS

 isolation

0.23

isol

0.19

 isol

0.18

 isolate

0.18

 isolated

0.16

 Subtract

0.12

 distrib

0.11

åŃ¤

0.11

istrib

0.11

emm

0.11

Activations Density 0.063%