ORPO: Monolithic Preference Optimization without Reference Model
Paper
•
2403.07691
•
Published
•
61
RLHF
-
-
sDPO: Don't Use Your Data All at Once
Paper • 2403.19270 • Published • 39 -
Teaching Large Language Models to Reason with Reinforcement Learning
Paper • 2403.04642 • Published • 46 -
Best Practices and Lessons Learned on Synthetic Data for Language Models
Paper • 2404.07503 • Published • 29 -
Rho-1: Not All Tokens Are What You Need
Paper • 2404.07965 • Published • 84 -
Learn Your Reference Model for Real Good Alignment
Paper • 2404.09656 • Published • 82 -
Dataset Reset Policy Optimization for RLHF
Paper • 2404.08495 • Published • 8 -
Insights into Alignment: Evaluating DPO and its Variants Across Multiple Tasks
Paper • 2404.14723 • Published • 10 -
RLHF Workflow: From Reward Modeling to Online RLHF
Paper • 2405.07863 • Published • 67 -
OpenRLHF: An Easy-to-use, Scalable and High-performance RLHF Framework
Paper • 2405.11143 • Published • 33 -
Mixtures of Experts Unlock Parameter Scaling for Deep RL
Paper • 2402.08609 • Published • 34 -
Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms
Paper • 2406.02900 • Published • 10 -
Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs
Paper • 2402.14740 • Published • 10 -
HelpSteer2: Open-source dataset for training top-performing reward models
Paper • 2406.08673 • Published • 16 -
Unpacking DPO and PPO: Disentangling Best Practices for Learning from Preference Feedback
Paper • 2406.09279 • Published • 1 -
Understanding the performance gap between online and offline alignment algorithms
Paper • 2405.08448 • Published • 14 -
Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision
Paper • 2312.09390 • Published • 32 -
Theoretical guarantees on the best-of-n alignment policy
Paper • 2401.01879 • Published -
Iterative Preference Learning from Human Feedback: Bridging Theory and Practice for RLHF under KL-Constraint
Paper • 2312.11456 • Published • 1 -
RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignment
Paper • 2304.06767 • Published • 2 -
Self-Play Preference Optimization for Language Model Alignment
Paper • 2405.00675 • Published • 24 -
Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs
Paper • 2406.10216 • Published • 2 -
Scaling Laws for Reward Model Overoptimization
Paper • 2210.10760 • Published -
AgentInstruct: Toward Generative Teaching with Agentic Flows
Paper • 2407.03502 • Published • 44 -
Online Merging Optimizers for Boosting Rewards and Mitigating Tax in Alignment
Paper • 2405.17931 • Published -
Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learning
Paper • 2405.00451 • Published -
Foundations of Reinforcement Learning and Interactive Decision Making
Paper • 2312.16730 • Published -
Agent Q: Advanced Reasoning and Learning for Autonomous AI Agents
Paper • 2408.07199 • Published • 20 -
Disentangling Length from Quality in Direct Preference Optimization
Paper • 2403.19159 • Published -
Imitating Language via Scalable Inverse Reinforcement Learning
Paper • 2409.01369 • Published -
Contrastive Prefence Learning: Learning from Human Feedback without RL
Paper • 2310.13639 • Published • 24 -
D2PO: Discriminator-Guided DPO with Response Evaluation Models
Paper • 2405.01511 • Published -
Anchored Preference Optimization and Contrastive Revisions: Addressing Underspecification in Alignment
Paper • 2408.06266 • Published • 9 -
Training Language Models to Self-Correct via Reinforcement Learning
Paper • 2409.12917 • Published • 133 -
The Perfect Blend: Redefining RLHF with Mixture of Judges
Paper • 2409.20370 • Published • 4 -
HelpSteer2-Preference: Complementing Ratings with Preferences
Paper • 2410.01257 • Published • 18 -
A Critical Evaluation of AI Feedback for Aligning Large Language Models
Paper • 2402.12366 • Published • 3 -
Rewarding Progress: Scaling Automated Process Verifiers for LLM Reasoning
Paper • 2410.08146 • Published