INDEX

Explanations

preservative, fixative, agent, inhibitor

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

agr

-0.10

lette

-0.10

/cms

-0.09

å¼ı

-0.09

-made

-0.09

ally

-0.09

mlink

-0.09

Ð¾Ð½Ð¾Ð²

-0.09

Dul

-0.09

SuccessListener

-0.09

POSITIVE LOGITS

 agent

0.22

agent

0.19

 agents

0.18

 Agent

0.17

Agent

0.15

åīĤ

0.14

_agent

0.14

 Agents

0.13

AGENT

0.13

-agent

0.13

Activations Density 0.065%