Ethical Guidance of LLMs: Revision history

19 December 2023

• curprev 20:3920:39, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,748 bytes +269‎ →‎Reinforcement Learning
• curprev 20:3220:32, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,479 bytes +11‎ →‎Reinforcement Learning
• curprev 20:3220:32, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,468 bytes +5‎ →‎Reinforcement Learning
• curprev 20:3120:31, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,463 bytes −2‎ →‎Reinforcement Learning
• curprev 20:3020:30, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,465 bytes +91‎ →‎Reinforcement Learning
• curprev 20:2820:28, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,374 bytes +1‎ →‎Reinforcement Learning
• curprev 20:2720:27, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,373 bytes −8‎ →‎Reinforcement Learning
• curprev 20:2720:27, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,381 bytes +46‎ →‎Reinforcement Learning
• curprev 20:2220:22, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 12,335 bytes +141‎ →‎Reinforcement Learning
• curprev 19:2919:29, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 12,194 bytes 0‎ →‎Limitations
• curprev 19:2919:29, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 12,194 bytes +117‎ →‎Overall
• curprev 19:0319:03, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 12,077 bytes +124‎ →‎Reinforcement Learning
• curprev 19:0219:02, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 11,953 bytes +323‎ →‎Limitations
• curprev 18:4218:42, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 11,630 bytes +219‎ →‎Preference/Reward Model
• curprev 18:1818:18, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 11,411 bytes +431‎ →‎Limitations
• curprev 17:4117:41, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 10,980 bytes −1‎ →‎Categorisation of Red Team Questions
• curprev 17:4117:41, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 10,981 bytes +358‎ →‎Categorisation of Red Team Questions
• curprev 17:2517:25, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 10,623 bytes +405‎ →‎Limitations
• curprev 16:5416:54, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 10,218 bytes +180‎ →‎Reinforcement Learning
• curprev 16:5016:50, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 10,038 bytes +33‎ →‎Reinforcement Learning
• curprev 16:4716:47, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 10,005 bytes +365‎ →‎Reinforcement Learning
• curprev 16:4616:46, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 9,640 bytes +78‎ →‎Deliverables
• curprev 16:2916:29, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 9,562 bytes +171‎ →‎Datasets
• curprev 16:1616:16, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 9,391 bytes +182‎ →‎Datasets
• curprev 16:1016:10, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 9,209 bytes +250‎ →‎Deliverables
• curprev 15:4315:43, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 8,959 bytes +43‎ →‎Introduction
• curprev 15:4315:43, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 8,916 bytes +106‎ →‎Reinforcement Learning
• curprev 14:4914:49, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 8,810 bytes +35‎ →‎Methodology
• curprev 14:4214:42, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 8,775 bytes +612‎ →‎Reinforcement Learning
• curprev 13:4713:47, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 8,163 bytes +645‎ →‎Reinforcement Learning
• curprev 13:0613:06, 19 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 7,518 bytes +211‎ →‎Supervised Learning
• curprev 10:4010:40, 19 December 2023‎ Carolina.maruganrubio talk contribs‎ 7,307 bytes +469‎ first attempt

14 December 2023

• curprev 13:2313:23, 14 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 6,838 bytes −1‎ →‎Data Preprocessing
• curprev 13:1613:16, 14 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 6,839 bytes +2,136‎ →‎Methodology

12 December 2023

• curprev 12:1912:19, 12 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 4,703 bytes +437‎ →‎Methodology

9 December 2023

• curprev 15:0715:07, 9 December 2023‎ Arundhati.balasubramaniam talk contribs‎ m 4,266 bytes +1,681‎ No edit summary
• curprev 14:3214:32, 9 December 2023‎ Arundhati.balasubramaniam talk contribs‎ 2,585 bytes +2,585‎ Created page with "== Abstract == Constitutional AI is a framework for creating artificial systems that can align with human values and preferences, without violating ethical principles. However, most existing methods for constitutional AI rely on human intervention, which can be costly, biased, and inconsistent. In this exploratory project, we replicate and extend the constitutional AI pipeline proposed by Anthropic, using Meta's Llama 2, a large language model with 7 billion parameters...."