Dynamic Normativity: Necessary and Sufficient Conditions for Value Alignment
Paper
•
2406.11039
•
Published
Aira
Aira is a series of chatbots developed as an experimentation playground for value alignment.
-
-
1
Aira Demo Portuguese
🤓Answer questions using a Portuguese language model
-
3
Aira Demo
🤓Chat with Aira, a research-focused assistant
-
nicholasKluge/Aira-2-portuguese-1B7
Text Generation • 2B • Updated • 43 • 2 -
nicholasKluge/Aira-2-1B5
Text Generation • 2B • Updated • 210 • 1 -
nicholasKluge/Aira-OPT-1B3
Text Generation • 1B • Updated • 27 -
nicholasKluge/Aira-2-1B1
Text Generation • 1B • Updated • 197 • 2 -
nicholasKluge/Aira-2-774M
Text Generation • 0.8B • Updated • 231 • 3 -
nicholasKluge/Aira-2-355M
Text Generation • 0.4B • Updated • 247 • 2 -
nicholasKluge/Aira-2-124M-DPO
Text Generation • 0.1B • Updated • 36 • 1 -
nicholasKluge/Aira-2-124M
Text Generation • 0.1B • Updated • 225 • 1 -
nicholasKluge/Aira-2-portuguese-560M
Text Generation • 0.6B • Updated • 52 • 2 -
nicholasKluge/Aira-2-portuguese-124M
Text Generation • 0.1B • Updated • 1.86k • 3 -
nicholasKluge/Aira-OPT-350M
Text Generation • 0.3B • Updated • 50 -
nicholasKluge/Aira-OPT-125M
Text Generation • 0.1B • Updated • 65 -
nicholasKluge/RewardModel
Text Classification • 0.1B • Updated • 157 • • 1 -
nicholasKluge/RewardModelPT
Text Classification • 0.1B • Updated • 78 -
nicholasKluge/Harmless-RewardModel
Text Classification • 0.1B • Updated • 97 • 1 -
nicholasKluge/Harmless-RewardModelPT
Text Classification • 0.1B • Updated • 21 -
nicholasKluge/ToxiGuardrail
Text Classification • 0.1B • Updated • 274k • 12 -
nicholasKluge/ToxiGuardrailPT
Text Classification • 0.1B • Updated • 1.72k • 1 -
nicholasKluge/instruct-aira-dataset-v3
Viewer • Updated • 100k • 65 • 6 -
nicholasKluge/instruct-aira-dataset-v2
Viewer • Updated • 163k • 66 • 4 -
nicholasKluge/instruct-aira-dataset
Viewer • Updated • 124k • 108 • 7 -
nicholasKluge/reward-aira-dataset
Viewer • Updated • 70k • 187 • 3 -
nicholasKluge/harmless-aira-dataset
Viewer • Updated • 16.6k • 215 • 4