CoRe: Co-Evolving Reward Models for Mitigating Latent Reward Hacking in Video Diffusion Models

arXiv:2609.36245 · cs.AI · Submitted 2026-09-28 · Read on arXiv

cs.AI

Submitted: 2026-09-28

Updated: 2026-09-28

Terminology

• Core — 1× in this paper · explained in 2 episodes
• Diffusion Model — 1× in this paper · explained in 9 episodes
• Diffusion Models — 1× in this paper · explained in 6 episodes
• Reward Hacking — 1× in this paper · explained in 8 episodes

Sources

• Off-Policy Corrected Reward Modeling for Reinforcement Learning from Human Feedback
• Generative Modeling via Drifting
• Reward Sharpness-Aware Fine-Tuning for Diffusion Models
• Flow Matching for Generative Modeling
• Decoupled Weight Decay Regularization
• Video Diffusion Alignment via Reward Gradients
• Wan: Open and Advanced Large-Scale Video Generative Models
• LiFT: Leveraging Human Feedback for Text-to-Video Model Alignment
• Reward Model Overoptimisation in Iterated RLHF
• DanceGRPO: Unleashing GRPO on Visual Generation
• Confronting Reward Overoptimization for Diffusion Models: A Perspective of Inductive and Primacy Biases
• VBench-2.0: Advancing Video Generation Benchmark Suite for Intrinsic Faithfulness

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems