paper-with-me

홈 › Papers

Objective Decoupling in Social Reinforcement Learning: Recovering Ground Truth from Sycophantic Majorities

2026-02-08 · Majid Ghasemi, Mark Crowley arxiv

Contemporary AI alignment strategies rely on a fragile premise: that human feedback, while noisy, remains a fundamentally truthful signal. In this paper, we identify this assumption as Dogma 4 of Reinforcement Learning (RL). We demonstrate that while this dogma holds in static environments, it fails in social settings where evaluators may be sycophantic, lazy, or adversarial. We prove that under Dogma 4, standard RL agents suffer from what we call Objective Decoupling, a structural failure mode where the agent's learned objective permanently separates from the latent ground truth, guaranteeing convergence to misalignment. To resolve this, we propose Epistemic Source Alignment (ESA). Unlike standard robust methods that rely on statistical consensus (trusting the majority), ESA utilizes sparse safety axioms to judge the source of the feedback rather than the signal itself. We prove that this "judging the judges" mechanism guarantees convergence to the true objective, even when a majority of evaluators are biased. Empirically, we show that while traditional consensus methods fail under majority collusion, our approach successfully recovers the optimal policy.

📄 PDF Abstract BibTeX arXiv:2602.08092

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

D$^2$NeRF: Self-Supervised Decoupling of Dynamic and Static Objects from a Monocular Video

2022-05-31 · Tianhao Wu, Fangcheng Zhong, Andrea Tagliasacchi, Forrester Cole 외

Given a monocular video, segmenting and decoupling dynamic objects while recovering the static environment is a widely studied problem in machine intelligence. Existing solutions usually approach this problem in the imag…

Image SegmentationNeRFSemantic SegmentationShadow Removal

Enhancing Deep Deterministic Policy Gradients on Continuous Control Tasks with Decoupled Prioritized Experience Replay

2025-12-04 · Mehmet Efe Lorasdagi, Dogan Can Cicek, Furkan Burak Mutlu, Suleyman Serdar Kozat arxiv

Background: Deep Deterministic Policy Gradient-based reinforcement learning algorithms utilize Actor-Critic architectures, where both networks are typically trained using identical batches of replayed transitions. Howeve…

Reinforcement LearningContinuous ControlOpenAI Gym

Learning Pareto-Optimal Rewards from Noisy Preferences: A Framework for Multi-Objective Inverse Reinforcement Learning

2025-05-17 · Kalyan Cherukuri, Aarav Lala

As generative agents become increasingly capable, alignment of their behavior with complex human values remains a fundamental challenge. Existing approaches often simplify human intent through reduction to a scalar rewar…

Perceptual Decoupling for Scalable Multi-modal Reasoning via Reward-Optimized Captioning

2025-06-05 · Yunhao Gou, Kai Chen, Zhili Liu, Lanqing Hong 외

Recent advances in slow-thinking language models (e.g., OpenAI-o1 and DeepSeek-R1) have demonstrated remarkable abilities in complex reasoning tasks by emulating human-like reflective cognition. However, extending such c…

MathVisual Grounding

On the Non-decoupling of Supervised Fine-tuning and Reinforcement Learning in Post-training

2026-01-12 · Xueyan Niu, Bo Bai, Wei Han, Weixi Zhang arxiv

Post-training of large language models routinely interleaves supervised fine-tuning (SFT) with reinforcement learning (RL). These two methods have different objectives: SFT minimizes the cross-entropy loss between model …

Reinforcement Learning