{"total":78,"rows":[{"id":"2503.21878","title":"Is Best-of-N the Best of Them? Coverage, Scaling, and Optimality in Inference-Time Alignment","created":"2025-03-27","cited_by_count":97,"area":"alignment"},{"id":"2502.12118","title":"Scaling Test-Time Compute Without Verification or RL is Suboptimal","created":"2025-02-17","cited_by_count":84,"area":"lr"},{"id":"2410.08847","title":"Unintentional Unalignment: Likelihood Displacement in Direct Preference Optimization","created":"2024-10-11","cited_by_count":67,"area":"alignment"},{"id":"2503.01067","title":"All Roads Lead to Likelihood: The Value of Reinforcement Learning in Fine-Tuning","created":"2025-03-03","cited_by_count":59,"area":"alignment"},{"id":"2503.11701","title":"A Survey of Direct Preference Optimization","created":"2025-03-12","cited_by_count":36,"area":"alignment"},{"id":"2506.19248","title":"Inference-Time Reward Hacking in Large Language Models","created":"2025-06-24","cited_by_count":34,"area":"alignment"},{"id":"2411.04625","title":"Sharp Analysis for KL-Regularized Contextual Bandits and RLHF","created":"2024-11-07","cited_by_count":33,"area":"alignment"},{"id":"2506.09477","title":"On a few pitfalls in KL divergence gradient estimation for RL","created":"2025-06-11","cited_by_count":29,"area":null},{"id":"2505.20686","title":"Accelerating RL for LLM Reasoning with Optimal Advantage Regression","created":"2025-05-27","cited_by_count":24,"area":"lr"},{"id":"2502.07460","title":"Logarithmic Regret for Online KL-Regularized Reinforcement Learning","created":"2025-02-11","cited_by_count":22,"area":"alignment"},{"id":"2505.23749","title":"Distortion of AI Alignment: Does Preference Optimization Optimize for Preferences?","created":"2025-05-29","cited_by_count":22,"area":"alignment"},{"id":"2502.01930","title":"Robust LLM Alignment via Distributionally Robust Direct Preference Optimization","created":"2025-02-04","cited_by_count":18,"area":"alignment"},{"id":"2602.01601","title":"Adaptive Rollout Allocation for Online Reinforcement Learning with Verifiable Rewards","created":"2026-02-02","cited_by_count":17,"area":null},{"id":"2503.08942","title":"Extragradient Preference Optimization (EGPO): Beyond Last-Iterate Convergence for Nash Learning from Human Feedback","created":"2025-03-11","cited_by_count":16,"area":"alignment"},{"id":"2511.07403","title":"SpatialThinker: Reinforcing Scene Graph-Grounded Spatial Reasoning via Dense Rewards","created":"2025-11-10","cited_by_count":16,"area":"lr"},{"id":"2412.09544","title":"Sail into the Headwind: Alignment via Robust Rewards and Dynamic Labels against Reward Hacking","created":"2024-12-12","cited_by_count":13,"area":"alignment"},{"id":"2601.10201","title":"Future-KL Regularized GRPO: Process-Level Credit Assignment from $f$-Divergence Regularization","created":"2026-01-15","cited_by_count":12,"area":"lr"},{"id":"2502.01203","title":"KL-Regularized RLHF with Multiple Reference Models: Exact Solutions and Sample Complexity","created":"2025-02-03","cited_by_count":11,"area":"alignment"},{"id":"2409.05622","title":"Forward KL Regularized Preference Optimization for Aligning Diffusion Policies","created":"2024-09-09","cited_by_count":10,"area":"diffusion"},{"id":"2502.06051","title":"Towards a Sharp Analysis of Offline Policy Learning for $f$-Divergence-Regularized Contextual Bandits","created":"2025-02-09","cited_by_count":10,"area":"offlinerl"},{"id":"2510.04088","title":"Offline Reinforcement Learning in Large State Spaces: Algorithms and Guarantees","created":"2025-10-05","cited_by_count":10,"area":"offlinerl"},{"id":"2410.20727","title":"Faster WIND: Accelerating Iterative Best-of-$N$ Distillation for LLM Alignment","created":"2024-10-28","cited_by_count":9,"area":"alignment"},{"id":"2505.19601","title":"Preference Optimization by Estimating the Ratio of the Data Distribution","created":"2025-05-26","cited_by_count":9,"area":"alignment"},{"id":"2503.06810","title":"Mitigating Preference Hacking in Policy Optimization with Pessimism","created":"2025-03-10","cited_by_count":8,"area":"alignment"},{"id":"2505.15694","title":"A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO","created":"2025-05-21","cited_by_count":8,"area":"alignment"},{"id":"2511.03317","title":"Diffusion-SDPO: Safeguarded Direct Preference Optimization for Diffusion Models","created":"2025-11-05","cited_by_count":7,"area":"diffusion"},{"id":"2503.00030","title":"RSPO: Regularized Self-Play Alignment of Large Language Models","created":"2025-02-24","cited_by_count":5,"area":"alignment"},{"id":"2505.20556","title":"Learning a Pessimistic Reward Model in RLHF","created":"2025-05-26","cited_by_count":5,"area":"alignment"},{"id":"2510.13694","title":"Information-Theoretic Reward Modeling for Stable RLHF: Detecting and Mitigating Reward Hacking","created":"2025-10-15","cited_by_count":5,"area":"alignment"},{"id":"2512.23816","title":"Improved Bounds for Private and Robust Alignment","created":"2025-12-29","cited_by_count":5,"area":"alignment"},{"id":"2602.23116","title":"Provably Efficient Regularized Online RLHF with Generalized Bilinear Preferences","created":"2026-02-26","cited_by_count":5,"area":"alignment"},{"id":"2502.06861","title":"Design Considerations in Offline Preference-based RL","created":"2025-02-08","cited_by_count":4,"area":"alignment"},{"id":"2502.19255","title":"Can RLHF be More Efficient with Imperfect Reward Models? A Policy Coverage Perspective","created":"2025-02-26","cited_by_count":4,"area":"alignment"},{"id":"2503.07426","title":"RePO: Understanding Preference Learning Through ReLU-Based Optimization","created":"2025-03-10","cited_by_count":4,"area":"alignment"},{"id":"2506.03757","title":"PPO in the Fisher-Rao geometry","created":"2025-06-04","cited_by_count":4,"area":"deeprl"},{"id":"2510.13512","title":"Offline and Online KL-Regularized RLHF under Differential Privacy","created":"2025-10-15","cited_by_count":4,"area":"alignment"},{"id":"2605.15417","title":"$f$-Trajectory Balance: A Loss Family for Tuning GFlowNets, Generative Models, and LLMs with Off- and On-Policy Data","created":"2026-05-14","cited_by_count":4,"area":null},{"id":"2502.10505","title":"Preference learning made easy: Everything should be understood through win rate","created":"2025-02-14","cited_by_count":3,"area":null},{"id":"2505.21395","title":"Square$\\chi$PO: Differentially Private and Robust $\\chi^2$-Preference Optimization in Offline Direct Alignment","created":"2025-05-27","cited_by_count":3,"area":"alignment"},{"id":"2506.01523","title":"Beyond RLHF: A Unified Theoretical Framework of Alignment","created":"2025-06-02","cited_by_count":3,"area":"alignment"},{"id":"2603.06797","title":"Best-of-Tails: Bridging Optimism and Pessimism in Inference-Time Alignment","created":"2026-03-06","cited_by_count":3,"area":"alignment"},{"id":"2605.09214","title":"Fast Rates for Offline Contextual Bandits with Forward-KL Regularization under Single-Policy Concentrability","created":"2026-05-09","cited_by_count":3,"area":null},{"id":"2606.30445","title":"When Does Online Imitation Learning Help in LLM Post-Training? The Role of (Non-)Realizability Beyond Horizon","created":"2026-06-29","cited_by_count":3,"area":"offlinerl"},{"id":"2411.06568","title":"Meta-Learning Objectives for Preference Optimization","created":"2024-11-10","cited_by_count":2,"area":"alignment"},{"id":"2502.17507","title":"C2-DPO: Constrained Controlled Direct Preference Optimization","created":"2025-02-22","cited_by_count":2,"area":"alignment"},{"id":"2509.19104","title":"Online Distributionally Robust LLM Alignment via Regression to Relative Reward","created":"2025-09-23","cited_by_count":2,"area":"alignment"},{"id":"2510.08256","title":"Mix- and MoE-DPO: A Variational Inference Approach to Direct Preference Optimization","created":"2025-10-09","cited_by_count":2,"area":"alignment"},{"id":"2512.16565","title":"Non-Asymptotic Global Convergence of PPO-Clip","created":"2025-12-18","cited_by_count":2,"area":null},{"id":"2512.21917","title":"Semiparametric Preference Optimization: Your Language Model is Secretly a Single-Index Model","created":"2025-12-26","cited_by_count":2,"area":"alignment"},{"id":"2602.00603","title":"Direct Preference Optimization with Rating Information: Practical Algorithms and Provable Gains","created":"2026-01-31","cited_by_count":2,"area":"alignment"},{"id":"2602.01685","title":"Semantic-aware Wasserstein Policy Regularization for Large Language Model Alignment","created":"2026-02-02","cited_by_count":2,"area":"alignment"},{"id":"2604.12086","title":"Robust Optimization for Mitigating Reward Hacking with Correlated Proxies","created":"2026-04-13","cited_by_count":2,"area":"alignment"},{"id":"2605.06977","title":"$f$-Divergence Regularized RLHF: Two Tales of Sampling and Unified Analyses","created":"2026-05-07","cited_by_count":2,"area":"alignment"},{"id":"2506.08681","title":"Mitigating Reward Over-optimization in Direct Alignment Algorithms with Importance Sampling","created":"2025-06-10","cited_by_count":1,"area":"alignment"},{"id":"2510.05526","title":"Provably Mitigating Corruption, Overoptimization, and Verbosity Simultaneously in Offline and Online RLHF/DPO Alignment","created":"2025-10-07","cited_by_count":1,"area":"alignment"},{"id":"2601.22083","title":"Latent Adversarial Regularization for Offline Preference Optimization","created":"2026-01-29","cited_by_count":1,"area":"alignment"},{"id":"2602.01357","title":"Your Self-Play Algorithm is Secretly an Adversarial Imitator: Understanding LLM Self-Play through the Lens of Imitation Learning","created":"2026-02-01","cited_by_count":1,"area":"synth"},{"id":"2605.03065","title":"OGPO: Sample Efficient Full-Finetuning of Generative Control Policies","created":"2026-05-04","cited_by_count":1,"area":null},{"id":"2605.06650","title":"Beyond Negative Rollouts: Positive-Only Policy Optimization with Implicit Negative Gradients","created":"2026-05-07","cited_by_count":1,"area":"lr"},{"id":"2605.30651","title":"LARK: Learnability-Grounded Trajectory Selection for Efficient Reasoning Distillation","created":"2026-05-28","cited_by_count":1,"area":"lr"},{"id":"2607.16240","title":"Normalized Rewards for Preference Optimization","created":"2026-06-26","cited_by_count":1,"area":"alignment"},{"id":"2609.12651","title":"Distortion of AI Alignment Revisited: RLHF is a Decent Utilitarian Aligner","created":"2026-09-11","cited_by_count":1,"area":"alignment"},{"id":"2507.07855","title":"DPO Unchained: Your Training Algorithm is Secretly Disentangled in Human Choice Theory (and its Loss' Convexity is Dispensable)","created":"2025-07-10","cited_by_count":0,"area":"alignment"},{"id":"2508.04225","title":"Symmetric Behavior Regularized Policy Optimization","created":"2025-08-06","cited_by_count":0,"area":"offlinerl"},{"id":"2509.20265","title":"Failure Modes of Maximum Entropy RLHF","created":"2025-09-24","cited_by_count":0,"area":"alignment"},{"id":"2601.19055","title":"Principled Fine-tuning of LLMs from User-Edits: A Medley of Preference, Supervision, and Reward","created":"2026-01-27","cited_by_count":0,"area":null},{"id":"2602.05933","title":"Approximation of Log-Partition Function in Policy Mirror Descent Induces Implicit Regularization for LLM Post-Training","created":"2026-02-05","cited_by_count":0,"area":null},{"id":"2602.06239","title":"Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution","created":"2026-02-05","cited_by_count":0,"area":"alignment"},{"id":"2602.06788","title":"Displacement-Resistant Extensions of DPO with Nonconvex $f$-Divergences","created":"2026-02-06","cited_by_count":0,"area":"alignment"},{"id":"2602.07738","title":"Learnable Chernoff Baselines for Inference-Time Alignment","created":"2026-02-08","cited_by_count":0,"area":null},{"id":"2604.10727","title":"Tail-Aware Information-Theoretic Bounds for LLM Alignment under Heavy-Tailed Rewards","created":"2026-04-12","cited_by_count":0,"area":"alignment"},{"id":"2605.24749","title":"How Neural Reward Models Learn Features for Policy Optimization: A Single-Index Analysis","created":"2026-05-23","cited_by_count":0,"area":"alignment"},{"id":"2606.01382","title":"Efficient Exploration for Iterative Nash Preference Optimization","created":"2026-05-31","cited_by_count":0,"area":"alignment"},{"id":"2606.04396","title":"Read the Trace, Steer the Path: Trajectory-Aware Reinforcement Learning for Diffusion Language Models","created":"2026-06-03","cited_by_count":0,"area":"diffusion"},{"id":"2606.22652","title":"A Markov Chain Approach to Preference Alignment","created":"2026-06-21","cited_by_count":0,"area":"alignment"},{"id":"2608.15402","title":"Towards a theory of inference-time alignment with unknown rewards","created":"2026-08-15","cited_by_count":0,"area":"alignment"},{"id":"2608.20011","title":"Manifold Drift in Flow Preference Optimization: A Root Cause of Reward Hacking","created":"2026-08-20","cited_by_count":0,"area":"diffusion"},{"id":"2610.01896","title":"Asynchronous LLM Post-Training: Group-Mass Capping and Convergence Analysis","created":"2026-10-01","cited_by_count":0,"area":"lr"}],"next_cursor":null}