INDEX

Explanations

training, obedience, commands, rewards

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 mèo

0.46

cleaning

0.43

𝕞

0.40

 stylesheets

0.39

 adoptar

0.39

kitty

0.39

рьох

0.38

ೃದ್ಧ

0.38

赤

0.38

catal

0.38

POSITIVE LOGITS

 Reward

0.60

 reward

0.59

 इनाम

0.57

 Instruct

0.57

 Obed

0.56

 obedience

0.55

 commands

0.55

 Commands

0.55

 Command

0.52

Reward

0.51

Activations Density 0.046%