{"total":135,"rows":[{"id":"2210.10760","title":"Scaling Laws for Reward Model Overoptimization","created":"2022-10-19","cited_by_count":1095,"area":"alignment"},{"id":"2307.15217","title":"Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback","created":"2023-07-27","cited_by_count":759,"area":"alignment"},{"id":"2309.11235","title":"OpenChat: Advancing Open-source Language Models with Mixed-Quality Data","created":"2023-09-20","cited_by_count":252,"area":null},{"id":"2302.08582","title":"Pretraining Language Models with Human Preferences","created":"2023-02-16","cited_by_count":247,"area":"alignment"},{"id":"2601.19897","title":"Self-Distillation Enables Continual Learning","created":"2026-01-27","cited_by_count":247,"area":"offlinerl"},{"id":"2501.06848","title":"A General Framework for Inference-time Scaling and Steering of Diffusion Models","created":"2025-01-12","cited_by_count":224,"area":"lr"},{"id":"2509.04259","title":"RL's Razor: Why Online Reinforcement Learning Forgets Less","created":"2025-09-04","cited_by_count":176,"area":"lr"},{"id":"2406.16838","title":"From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models","created":"2024-06-24","cited_by_count":137,"area":null},{"id":"2310.17022","title":"Controlled Decoding from Language Models","created":"2023-10-25","cited_by_count":135,"area":null},{"id":"2303.16755","title":"Training Language Models with Language Feedback at Scale","created":"2023-03-28","cited_by_count":128,"area":"alignment"},{"id":"2401.01879","title":"Theoretical guarantees on the best-of-n alignment policy","created":"2024-01-03","cited_by_count":124,"area":"lr"},{"id":"2302.08215","title":"Aligning Language Models with Preferences through f-divergence Minimization","created":"2023-02-16","cited_by_count":120,"area":"alignment"},{"id":"2404.08555","title":"RLHF Deciphered: A Critical Analysis of Reinforcement Learning from Human Feedback for LLMs","created":"2024-04-12","cited_by_count":90,"area":"alignment"},{"id":"2404.17546","title":"Probabilistic Inference in Language Models via Twisted Sequential Monte Carlo","created":"2024-04-26","cited_by_count":90,"area":"alignment"},{"id":"2206.00761","title":"On Reinforcement Learning and Distribution Matching for Fine-Tuning Language Models with no Catastrophic Forgetting","created":"2022-06-01","cited_by_count":82,"area":null},{"id":"2306.03081","title":"Sequential Monte Carlo Steering of Large Language Models using Probabilistic Programs","created":"2023-06-05","cited_by_count":80,"area":null},{"id":"2310.00035","title":"LoRA ensembles for large language model fine-tuning","created":"2023-09-29","cited_by_count":67,"area":null},{"id":"2402.02992","title":"Decoding-time Realignment of Language Models","created":"2024-02-05","cited_by_count":67,"area":"alignment"},{"id":"2405.19316","title":"Robust Preference Optimization through Reward Model Distillation","created":"2024-05-29","cited_by_count":67,"area":"alignment"},{"id":"2310.12962","title":"An Emulator for Fine-Tuning Large Language Models using Small Language Models","created":"2023-10-19","cited_by_count":61,"area":null},{"id":"2402.07314","title":"Online Iterative Reinforcement Learning from Human Feedback with General Preference Model","created":"2024-02-11","cited_by_count":52,"area":"alignment"},{"id":"2510.17516","title":"SimBench: Benchmarking the Ability of Large Language Models to Simulate Human Behaviors","created":"2025-10-20","cited_by_count":52,"area":null},{"id":"2504.12522","title":"Evaluating the Diversity and Quality of LLM Generated Content","created":"2025-04-16","cited_by_count":51,"area":"lr"},{"id":"2403.03185","title":"Correlated Proxies: A New Definition and Improved Mitigation for Reward Hacking","created":"2024-03-05","cited_by_count":50,"area":"alignment"},{"id":"2405.06639","title":"Value Augmented Sampling for Language Model Alignment and Personalization","created":"2024-05-10","cited_by_count":49,"area":"alignment"},{"id":"2405.19534","title":"Preference Learning Algorithms Do Not Learn Preference Rankings","created":"2024-05-29","cited_by_count":48,"area":"alignment"},{"id":"2404.01730","title":"Asymptotics of Language Model Alignment","created":"2024-04-02","cited_by_count":47,"area":"alignment"},{"id":"2504.13139","title":"Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlo","created":"2025-04-17","cited_by_count":46,"area":null},{"id":"2407.00617","title":"Iterative Nash Policy Optimization: Aligning LLMs with General Preferences via No-Regret Learning","created":"2024-06-30","cited_by_count":45,"area":"alignment"},{"id":"2402.00856","title":"Towards Efficient Exact Optimization of Language Model Alignment","created":"2024-02-01","cited_by_count":37,"area":"alignment"},{"id":"2412.02980","title":"Surveying the Effects of Quality, Diversity, and Complexity in Synthetic Data From Large Language Models","created":"2024-12-04","cited_by_count":35,"area":"synth"},{"id":"2412.19792","title":"InfAlign: Inference-aware language model alignment","created":"2024-12-27","cited_by_count":35,"area":"alignment"},{"id":"2305.15065","title":"Inference-Time Policy Adapters (IPA): Tailoring Extreme-Scale LMs without Fine-tuning","created":"2023-05-24","cited_by_count":33,"area":null},{"id":"2211.08714","title":"Reward Gaming in Conditional Text Generation","created":"2022-11-16","cited_by_count":30,"area":null},{"id":"2410.07927","title":"Efficient Reinforcement Learning with Large Language Model Priors","created":"2024-10-10","cited_by_count":30,"area":null},{"id":"2503.18929","title":"Trajectory Balance with Asynchrony: Decoupling Exploration and Learning for Fast, Scalable LLM Post-Training","created":"2025-03-24","cited_by_count":29,"area":"lr"},{"id":"2401.13586","title":"Instruction Fine-Tuning: Does Prompt Loss Matter?","created":"2024-01-24","cited_by_count":26,"area":null},{"id":"2402.00742","title":"Transforming and Combining Rewards for Aligning Large Language Models","created":"2024-02-01","cited_by_count":25,"area":"alignment"},{"id":"2502.16852","title":"Improving LLM General Preference Alignment via Optimistic Online Mirror Descent","created":"2025-02-24","cited_by_count":24,"area":"alignment"},{"id":"2309.09055","title":"Exploring the impact of low-rank adaptation on the performance, efficiency, and regularization of RLHF","created":"2023-09-16","cited_by_count":20,"area":"alignment"},{"id":"2502.16320","title":"Direct Alignment with Heterogeneous Preferences","created":"2025-02-22","cited_by_count":19,"area":"alignment"},{"id":"2512.21852","title":"A Comedy of Estimators: On KL Regularization in RL Training of LLMs","created":"2025-12-26","cited_by_count":18,"area":"lr"},{"id":"2502.06999","title":"Outsourced diffusion sampling: Efficient posterior inference in latent spaces of generative models","created":"2025-02-10","cited_by_count":17,"area":"diffusion"},{"id":"2406.00049","title":"QUEST: Quality-Aware Metropolis-Hastings Sampling for Machine Translation","created":"2024-05-28","cited_by_count":16,"area":null},{"id":"2507.18537","title":"TTS-VAR: A Test-Time Scaling Framework for Visual Auto-Regressive Generation","created":"2025-07-24","cited_by_count":15,"area":"lr"},{"id":"2502.00968","title":"CoDe: Blockwise Control for Denoising Diffusion Models","created":"2025-02-03","cited_by_count":14,"area":"diffusion"},{"id":"2502.14289","title":"Drift: Decoding-time Personalized Alignments with Implicit User Preferences","created":"2025-02-20","cited_by_count":13,"area":"alignment"},{"id":"2504.02367","title":"Reinforcement Fine-Tuning for Materials Design","created":"2025-04-03","cited_by_count":13,"area":null},{"id":"2302.00805","title":"Conditioning Predictive Models: Risks and Strategies","created":"2023-02-02","cited_by_count":11,"area":null},{"id":"2506.04118","title":"Guided Speculative Inference for Efficient Test-Time Alignment of LLMs","created":"2025-06-04","cited_by_count":11,"area":"alignment"},{"id":"2603.11682","title":"Entropy-Preserving Reinforcement Learning","created":"2026-03-12","cited_by_count":11,"area":"lr"},{"id":"2305.17760","title":"Language Models are Bounded Pragmatic Speakers: Understanding RLHF from a Bayesian Cognitive Modeling Perspective","created":"2023-05-28","cited_by_count":10,"area":"alignment"},{"id":"2310.16763","title":"SuperHF: Supervised Iterative Learning from Human Feedback","created":"2023-10-25","cited_by_count":10,"area":"alignment"},{"id":"2410.22296","title":"Generalists vs. Specialists: Evaluating LLMs on Highly-Constrained Biophysical Sequence Optimization Tasks","created":"2024-10-29","cited_by_count":10,"area":null},{"id":"2606.26091","title":"On-Policy Self-Distillation with Sampled Demonstrations Reduces Output Diversity","created":"2026-06-24","cited_by_count":10,"area":"distill"},{"id":"2412.13862","title":"Energy-Based Preference Model Offers Better Offline Alignment than the Bradley-Terry Preference Model","created":"2024-12-18","cited_by_count":9,"area":"alignment"},{"id":"2605.11613","title":"From Generic Correlation to Input-Specific Credit in On-Policy Self Distillation","created":"2026-05-12","cited_by_count":9,"area":"distill"},{"id":"2510.09714","title":"All Code, No Thought: Current Language Models Struggle to Reason in Ciphered Language","created":"2025-10-10","cited_by_count":8,"area":"lr"},{"id":"2512.15605","title":"Autoregressive Language Models are Secretly Energy-Based Models: Insights into the Lookahead Capabilities of Next-Token Prediction","created":"2025-12-17","cited_by_count":8,"area":null},{"id":"2310.02263","title":"Automatic Pair Construction for Contrastive Post-training","created":"2023-10-03","cited_by_count":7,"area":null},{"id":"2407.15229","title":"A Practical Analysis of Human Alignment with *PO","created":"2024-07-21","cited_by_count":7,"area":"alignment"},{"id":"2504.18026","title":"Addressing Concept Mislabeling in Concept Bottleneck Models Through Preference Optimization","created":"2025-04-25","cited_by_count":7,"area":"alignment"},{"id":"2507.08390","title":"Inference-Time Scaling of Diffusion Language Models via Trajectory Refinement","created":"2025-07-11","cited_by_count":7,"area":"lr"},{"id":"2602.03445","title":"CRL-VLA: Continual Vision-Language-Action Learning","created":"2026-02-03","cited_by_count":7,"area":"vla"},{"id":"2402.03284","title":"Deal, or no deal (or who knows)? Forecasting Uncertainty in Conversations using Large Language Models","created":"2024-02-05","cited_by_count":6,"area":null},{"id":"2601.00898","title":"Dichotomous Diffusion Policy Optimization","created":"2025-12-31","cited_by_count":6,"area":"diffusion"},{"id":"2312.01957","title":"Distilled Self-Critique of LLMs with Synthetic Data: a Bayesian Perspective","created":"2023-12-04","cited_by_count":5,"area":"alignment"},{"id":"2410.00362","title":"FedPT: Federated Proxy-Tuning of Large Language Models on Resource-Constrained Edge Devices","created":"2024-10-01","cited_by_count":5,"area":null},{"id":"2509.01632","title":"Relative Trajectory Balance is equivalent to Trust-PCL","created":"2025-09-01","cited_by_count":5,"area":"diffusion"},{"id":"2510.21978","title":"Beyond Reasoning Gains: Mitigating General-Capability Forgetting in Large Reasoning Models","created":"2025-10-24","cited_by_count":5,"area":"lr"},{"id":"2602.16570","title":"Steering diffusion models with quadratic rewards: a fine-grained analysis","created":"2026-02-18","cited_by_count":5,"area":"diffusion"},{"id":"2604.14265","title":"Reinforcement Learning via Value Gradient Flow","created":"2026-04-15","cited_by_count":5,"area":"offlinerl"},{"id":"2605.23346","title":"Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion","created":"2026-05-22","cited_by_count":5,"area":"diffusion"},{"id":"2505.01954","title":"Semantic Probabilistic Control of Language Models","created":"2025-05-04","cited_by_count":4,"area":null},{"id":"2505.17795","title":"DialogXpert: Driving Intelligent and Emotion-Aware Conversations through Online Value-Based Reinforcement Learning with LLM Priors","created":"2025-05-23","cited_by_count":4,"area":null},{"id":"2510.00502","title":"Diffusion Alignment as Variational Expectation-Maximization","created":"2025-10-01","cited_by_count":4,"area":"diffusion"},{"id":"2603.17218","title":"Alignment Makes Language Models Normative, Not Descriptive","created":"2026-03-17","cited_by_count":4,"area":"alignment"},{"id":"2406.11039","title":"Dynamic Normativity: Necessary and Sufficient Conditions for Value Alignment","created":"2024-06-16","cited_by_count":3,"area":"alignment"},{"id":"2412.02626","title":"Time-Reversal Provides Unsupervised Feedback to LLMs","created":"2024-12-03","cited_by_count":3,"area":null},{"id":"2502.16433","title":"Sequence-level Large Language Model Training with Contrastive Preference Optimization","created":"2025-02-23","cited_by_count":3,"area":"alignment"},{"id":"2503.15129","title":"Aligning Crowd-sourced Human Feedback for Reinforcement Learning on Code Generation by Large Language Models","created":"2025-03-19","cited_by_count":3,"area":"agents"},{"id":"2510.21296","title":"An Evidence-Based Post-Hoc Adjustment Framework for Anomaly Detection Under Data Contamination","created":"2025-10-24","cited_by_count":3,"area":null},{"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":"2605.26552","title":"Aligning Few-Step Generative Models by Amortizing Sample-based Variational Inference","created":"2026-05-26","cited_by_count":3,"area":null},{"id":"2606.29604","title":"Mechanistically Eliciting Latent Behaviors in Language Models","created":"2026-06-28","cited_by_count":3,"area":null},{"id":"2410.00508","title":"FlipGuard: Defending Preference Alignment against Update Regression with Constrained Optimization","created":"2024-10-01","cited_by_count":2,"area":"alignment"},{"id":"2410.06213","title":"RL, but don't do anything I wouldn't do","created":"2024-10-08","cited_by_count":2,"area":"deeprl"},{"id":"2502.12776","title":"Portable Reward Tuning: Towards Reusable Fine-Tuning across Different Pretrained Models","created":"2025-02-18","cited_by_count":2,"area":"lr"},{"id":"2505.08971","title":"Prioritizing Image-Related Tokens Enhances Vision-Language Pre-Training","created":"2025-05-13","cited_by_count":2,"area":"mllm"},{"id":"2507.02315","title":"Improving Constrained Language Generation via Self-Distilled Twisted Sequential Monte Carlo","created":"2025-07-03","cited_by_count":2,"area":null},{"id":"2402.05829","title":"Limitations of Agents Simulated by Predictive Models","created":"2024-02-08","cited_by_count":1,"area":"agents"},{"id":"2406.13660","title":"Towards Minimal Targeted Updates of Language Models with Targeted Negative Training","created":"2024-06-19","cited_by_count":1,"area":null},{"id":"2505.23761","title":"Differential Information Distribution: A Bayesian Perspective on Direct Preference Optimization","created":"2025-05-29","cited_by_count":1,"area":"alignment"},{"id":"2506.09416","title":"Noise Conditional Variational Score Distillation","created":"2025-06-11","cited_by_count":1,"area":"diffusion"},{"id":"2508.10409","title":"Dataset Construction for Training LLM to Learn Analog Circuit Knowledge","created":"2025-08-14","cited_by_count":1,"area":null},{"id":"2509.23667","title":"Why Alignment Must Precede Distillation: A Minimal Working Explanation","created":"2025-09-28","cited_by_count":1,"area":"alignment"},{"id":"2510.25897","title":"MIRO: MultI-Reward cOnditioned pretraining improves T2I quality and efficiency","created":"2025-10-29","cited_by_count":1,"area":"alignment"},{"id":"2601.20198","title":"DeRaDiff: Denoising Time Realignment of Diffusion Models","created":"2026-01-28","cited_by_count":1,"area":"diffusion"},{"id":"2604.17290","title":"Probabilistic Programs of Thought","created":"2026-04-19","cited_by_count":1,"area":"lr"},{"id":"2604.28036","title":"Exponential families from a single KL identity","created":"2026-04-30","cited_by_count":1,"area":null}],"next_cursor":"100"}