Deep Reinforcement Learning from Human Preferences

N/A

# link_1v5y0rmw15d.pdf ## Metadata - PDFFormatVersion=1.3 - IsLinearized=false - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - Subject=Neural Information Processing Systems http://nips.cc/ - Custom.Publisher=Curran Associates, Inc. - Custom.Language=en-US - Custom.Created=2017 - Custom.EventType=Poster - Custom.Description-Abstract=For sophisticated reinforcement learning (RL) systems to interact usefully with real-world environments, we need to communicate complex goals to these systems. In this work, we explore goals defined in terms o

saved by

• 6824

related reading

• Reinforcement learning from human feedback - Wikipediaen.wikipedia.org
• Learning through human feedback — Google DeepMinddeepmind.google
• RLHF: Reinforcement Learning from Human Feedbackhuyenchip.com
• RLHF Bookrlhfbook.com
• Reward Hacking in Reinforcement Learning | Lil'Loglilianweng.github.io
• [2305.18290] Direct Preference Optimization: Your Language Model is Secretly a Reward Modelarxiv.org
• The Era of Experience Paper.pdfstorage.googleapis.com
• Reward is not the optimization target — LessWronglesswrong.com
• Deep Reinforcement Learning Doesn't Work Yetalexirpan.com
• Part 1: Key Concepts in RL - Spinning Up documentationspinningup.openai.com
• [2309.00267] RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedbackarxiv.org
• rlhfbook.com/book.pdfrlhfbook.com