INDEX

Explanations

prepositions and articles

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

 strav

-0.14

 pract

-0.12

 pornÃ´

-0.11

 follando

-0.11

 observ

-0.10

 ancor

-0.10

observ

-0.10

Ov

-0.10

 jadx

-0.10

RuleContext

-0.10

POSITIVE LOGITS

la

0.16

il

0.15

un

0.12

le

0.12

aff

0.12

 nond

0.12

 accord

0.12

Activations Density 0.128%