{"total":76,"rows":[{"id":"2405.14734","title":"SimPO: Simple Preference Optimization with a Reference-Free Reward","created":"2024-05-23","cited_by_count":963,"area":"alignment"},{"id":"2506.01937","title":"RewardBench 2: Advancing Reward Model Evaluation","created":"2025-06-02","cited_by_count":150,"area":"alignment"},{"id":"2507.18624","title":"Checklists Are Better Than Reward Models For Aligning Language Models","created":"2025-07-24","cited_by_count":128,"area":"alignment"},{"id":"2411.01111","title":"Rule Based Rewards for Language Model Safety","created":"2024-11-02","cited_by_count":108,"area":"alignment"},{"id":"2411.17451","title":"VL-RewardBench: A Challenging Benchmark for Vision-Language Generative Reward Models","created":"2024-11-26","cited_by_count":86,"area":"alignment"},{"id":"2408.02666","title":"Self-Taught Evaluators","created":"2024-08-05","cited_by_count":72,"area":"alignment"},{"id":"2410.09893","title":"RMB: Comprehensively Benchmarking Reward Models in LLM Alignment","created":"2024-10-13","cited_by_count":61,"area":"alignment"},{"id":"2410.14632","title":"Diverging Preferences: When do Annotators Disagree and do Models Know?","created":"2024-10-18","cited_by_count":47,"area":"alignment"},{"id":"2409.00598","title":"Automatic Pseudo-Harmful Prompt Generation for Evaluating False Refusals in Large Language Models","created":"2024-09-01","cited_by_count":35,"area":"alignment"},{"id":"2409.15268","title":"Style Outweighs Substance: Failure Modes of LLM Judges in Alignment Benchmarking","created":"2024-09-23","cited_by_count":35,"area":"alignment"},{"id":"2408.15549","title":"WildFeedback: Aligning LLMs With In-situ User Interactions And Feedback","created":"2024-08-28","cited_by_count":31,"area":"alignment"},{"id":"2510.17314","title":"Auto-Rubric: Learning From Implicit Weights to Explicit Rubrics for Reward Modeling","created":"2025-10-20","cited_by_count":30,"area":"alignment"},{"id":"2506.05598","title":"SynthesizeMe! Inducing Persona-Guided Prompts for Personalized Reward Models in LLMs","created":"2025-06-05","cited_by_count":29,"area":"alignment"},{"id":"2603.12273","title":"Aligning Language Models from User Interactions","created":"2026-02-18","cited_by_count":28,"area":"alignment"},{"id":"2511.01689","title":"Open Character Training: Shaping the Persona of AI Assistants through Constitutional AI","created":"2025-11-03","cited_by_count":24,"area":"alignment"},{"id":"2409.11564","title":"Preference Tuning with Human Feedback on Language, Speech, and Vision Tasks: A Survey","created":"2024-09-17","cited_by_count":23,"area":"alignment"},{"id":"2502.11555","title":"Equilibrate RLHF: Towards Balancing Helpfulness-Safety Trade-off in Large Language Models","created":"2025-02-17","cited_by_count":22,"area":"alignment"},{"id":"2503.18991","title":"Inverse Reinforcement Learning with Dynamic Reward Scaling for LLM Alignment","created":"2025-03-23","cited_by_count":20,"area":"alignment"},{"id":"2406.10630","title":"Emerging Safety Attack and Defense in Federated Instruction Tuning of Large Language Models","created":"2024-06-15","cited_by_count":18,"area":"alignment"},{"id":"2503.02574","title":"LLM-Safety Evaluations Lack Robustness","created":"2025-03-04","cited_by_count":18,"area":"alignment"},{"id":"2505.16869","title":"MPO: Multilingual Safety Alignment via Reward Gap Optimization","created":"2025-05-22","cited_by_count":18,"area":"alignment"},{"id":"2502.04322","title":"Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions","created":"2025-02-06","cited_by_count":15,"area":"alignment"},{"id":"2407.10944","title":"Naturally Occurring Feedback is Common, Extractable and Useful","created":"2024-07-15","cited_by_count":14,"area":"alignment"},{"id":"2505.05465","title":"ComPO: Preference Alignment via Comparison Oracles","created":"2025-05-08","cited_by_count":13,"area":"alignment"},{"id":"2506.03690","title":"Robust Preference Optimization via Dynamic Target Margins","created":"2025-06-04","cited_by_count":13,"area":"alignment"},{"id":"2408.16961","title":"The Future of Open Human Feedback","created":"2024-08-15","cited_by_count":10,"area":"alignment"},{"id":"2509.01301","title":"Culture is Everywhere: A Call for Intentionally Cultural Evaluation","created":"2025-09-01","cited_by_count":10,"area":"alignment"},{"id":"2510.26707","title":"Value Drifts: Tracing Value Alignment During LLM Post-Training","created":"2025-10-30","cited_by_count":10,"area":"alignment"},{"id":"2504.03612","title":"AIR: A Systematic Analysis of Annotations, Instructions, and Response Pairs in Preference Dataset","created":"2025-04-04","cited_by_count":9,"area":"alignment"},{"id":"2406.16783","title":"M2Lingual: Enhancing Multilingual, Multi-Turn Instruction Alignment in Large Language Models","created":"2024-06-24","cited_by_count":7,"area":"alignment"},{"id":"2410.01735","title":"LASeR: Learning to Adaptively Select Reward Models with Multi-Armed Bandits","created":"2024-10-02","cited_by_count":7,"area":"alignment"},{"id":"2505.02363","title":"SIMPLEMIX: Frustratingly Simple Mixing of Off- and On-policy Data in Language Model Preference Learning","created":"2025-05-05","cited_by_count":6,"area":"alignment"},{"id":"2506.17353","title":"Differentiation-Based Extraction of Proprietary Data from Fine-Tuned LLMs","created":"2025-06-20","cited_by_count":6,"area":"alignment"},{"id":"2510.24235","title":"PaTaRM: Bridging Pairwise and Pointwise Signals via Preference-Aware Task-Adaptive Reward Modeling","created":"2025-10-28","cited_by_count":6,"area":"alignment"},{"id":"2602.07892","title":"Safety Alignment as Continual Learning: Mitigating the Alignment Tax via Orthogonal Gradient Projection","created":"2026-02-08","cited_by_count":6,"area":"alignment"},{"id":"2506.03968","title":"From Real to Synthetic: Synthesizing Millions of Diversified and Complicated User Instructions with Attributed Grounding","created":"2025-06-04","cited_by_count":5,"area":"alignment"},{"id":"2506.04180","title":"SuperWriter: Reflection-Driven Long-Form Generation with Large Language Models","created":"2025-06-04","cited_by_count":5,"area":"alignment"},{"id":"2501.18511","title":"WILDCHAT-50M: A Deep Dive Into the Role of Synthetic Data in Post-Training","created":"2025-01-30","cited_by_count":4,"area":"alignment"},{"id":"2602.15222","title":"Automatically Finding Reward Model Biases","created":"2026-02-16","cited_by_count":4,"area":"alignment"},{"id":"2510.06370","title":"EVALUESTEER: Measuring Reward Model Steerability Towards Values and Preferences","created":"2025-10-07","cited_by_count":3,"area":"alignment"},{"id":"2510.18081","title":"Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth","created":"2025-10-20","cited_by_count":3,"area":"alignment"},{"id":"2601.03589","title":"OLA: Output Language Alignment in Code-Switched LLM Interactions","created":"2026-01-07","cited_by_count":3,"area":"alignment"},{"id":"2602.08829","title":"WildReward: Learning Reward Models from In-the-Wild Human Interactions","created":"2026-02-09","cited_by_count":3,"area":"alignment"},{"id":"2603.03291","title":"One Bias After Another: Mechanistic Reward Shaping and Persistent Biases in Language Reward Models","created":"2026-02-06","cited_by_count":3,"area":"alignment"},{"id":"2604.02686","title":"Beyond Semantic Manipulation: Token-Space Attacks on Reward Models","created":"2026-04-03","cited_by_count":3,"area":"alignment"},{"id":"2606.21795","title":"Discretizing Reward Models","created":"2026-06-19","cited_by_count":3,"area":"alignment"},{"id":"2607.01830","title":"Many Voices, One Reward: Multi-Role Rubric Generation for LLM Judging and Reward Modeling","created":"2026-07-02","cited_by_count":3,"area":"alignment"},{"id":"2503.08035","title":"Group Preference Alignment: Customized LLM Response Generation from In-Situ Conversations","created":"2025-03-11","cited_by_count":2,"area":"alignment"},{"id":"2511.06512","title":"EASE: Practical and Efficient Safety Alignment for Small Language Models","created":"2025-11-09","cited_by_count":2,"area":"alignment"},{"id":"2602.09653","title":"ClinAlign: Scaling Healthcare Alignment from Clinician Preference","created":"2026-02-10","cited_by_count":2,"area":"alignment"},{"id":"2602.13551","title":"Small Reward Models via Backward Inference","created":"2026-02-14","cited_by_count":2,"area":"alignment"},{"id":"2603.12795","title":"SteerRM: Debiasing Reward Models via Sparse Autoencoders","created":"2026-03-13","cited_by_count":2,"area":"alignment"},{"id":"2603.12963","title":"Long-form RewardBench: Evaluating Reward Models for Long-form Generation","created":"2026-03-13","cited_by_count":2,"area":"alignment"},{"id":"2604.13828","title":"MUSE: Multi-Domain Chinese User Simulation via Self-Evolving Profiles and Rubric-Guided Alignment","created":"2026-04-15","cited_by_count":2,"area":"alignment"},{"id":"2406.16377","title":"On the Transformations across Reward Model, Parameter Update, and In-Context Prompt","created":"2024-06-24","cited_by_count":1,"area":"alignment"},{"id":"2502.19463","title":"Hedging and Non-Affirmation: Quantifying LLM Alignment on Questions of Human Rights","created":"2025-02-26","cited_by_count":1,"area":"alignment"},{"id":"2510.06670","title":"PIKA: Expert-Level Synthetic Datasets for Post-Training Alignment from Scratch","created":"2025-10-08","cited_by_count":1,"area":"alignment"},{"id":"2510.15514","title":"Voting with the Graph: Stable RLAIF via Topological Consistency Maximization","created":"2025-10-17","cited_by_count":1,"area":"alignment"},{"id":"2511.10985","title":"When Data is the Algorithm: A Systematic Study and Curation of Preference Optimization Datasets","created":"2025-11-14","cited_by_count":1,"area":"alignment"},{"id":"2607.14240","title":"Align AI to Dynamic Human-AI Workflows","created":"2026-07-15","cited_by_count":1,"area":"alignment"},{"id":"2610.05308","title":"RubricArmor: Adversarial Evolution Improves LLM-Based Rubric Generation","created":"2026-10-04","cited_by_count":1,"area":"alignment"},{"id":"2507.15281","title":"A Novel Self-Evolution Framework for Large Language Models","created":"2025-07-21","cited_by_count":0,"area":"alignment"},{"id":"2508.09937","title":"A Comprehensive Evaluation framework of Alignment Techniques for LLMs","created":"2025-08-13","cited_by_count":0,"area":"alignment"},{"id":"2509.26305","title":"Feedback Forensics: A Toolkit to Measure AI Personality","created":"2025-09-30","cited_by_count":0,"area":"alignment"},{"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":"2604.27453","title":"From Coarse to Fine: Benchmarking and Reward Modeling for Writing-Centric Generation Tasks","created":"2026-04-30","cited_by_count":0,"area":"alignment"},{"id":"2605.00022","title":"Putting HUMANS first: Efficient LAM Evaluation with Human Preference Alignment","created":"2026-04-20","cited_by_count":0,"area":"alignment"},{"id":"2605.04029","title":"Stayin' Aligned Over Time: Towards Longitudinal Human-LLM Alignment via Contextual Reflection and Privacy-Preserving Behavioral Data","created":"2026-05-05","cited_by_count":0,"area":"alignment"},{"id":"2605.19966","title":"Detecting Fluent Optimization-Based Adversarial Prompts via Sequential Entropy Changes","created":"2026-05-19","cited_by_count":0,"area":"alignment"},{"id":"2605.20654","title":"REFLECTOR: Internalizing Step-wise Reflection against Indirect Jailbreak","created":"2026-05-20","cited_by_count":0,"area":"alignment"},{"id":"2607.02914","title":"Oyster-II: Reinforcement Learning for Constructive Safety Alignment in Large Language Models","created":"2026-07-03","cited_by_count":0,"area":"alignment"},{"id":"2607.20485","title":"Expectation Alignment of Language Models for Real-World User Expectations","created":"2026-06-02","cited_by_count":0,"area":"alignment"},{"id":"2608.05409","title":"Mood Matters: How Syntactic Sensitivity Undermines Safety Alignment","created":"2026-08-05","cited_by_count":0,"area":"alignment"},{"id":"2609.01455","title":"When Safety Routing Breaks: Understanding Alignment Fragility under Benign Fine-Tuning","created":"2026-09-01","cited_by_count":0,"area":"alignment"},{"id":"2609.34519","title":"EOPSA: Efficient On-Policy Self-Distilled Safety Alignment","created":"2026-09-28","cited_by_count":0,"area":"alignment"},{"id":"2609.36849","title":"Does the Unsafe Gradient Survive a Conversation? On the Fragility of Gradient-Based Jailbreak Detection in Multi-Turn Dialogue","created":"2026-09-29","cited_by_count":0,"area":"alignment"}],"next_cursor":null}