PIRF: Physics-Informed Reward Fine-Tuning for Diffusion Models

• URL: http://arxiv.org/abs/2509.20570v1
• Date: Wed, 24 Sep 2025 21:23:03 GMT
• Title: PIRF: Physics-Informed Reward Fine-Tuning for Diffusion Models
• Abstract summary: We frame physics-informed generation as a sparse reward optimization problem, where adherence to physical constraints is treated as a reward signal.<n>We introduce Physics-Informed Reward Fine-tuning (PIRF), a method that bypasses value approximation by computing trajectory-level rewards and backpropagating their gradients directly.<n>PIRF consistently achieves superior physical enforcement under efficient sampling regimes, highlighting the potential of reward fine-tuning for advancing scientific generative modeling.
• Score: 11.791955441600825
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Diffusion models have demonstrated strong generative capabilities across scientific domains, but often produce outputs that violate physical laws. We propose a new perspective by framing physics-informed generation as a sparse reward optimization problem, where adherence to physical constraints is treated as a reward signal. This formulation unifies prior approaches under a reward-based paradigm and reveals a shared bottleneck: reliance on diffusion posterior sampling (DPS)-style value function approximations, which introduce non-negligible errors and lead to training instability and inference inefficiency. To overcome this, we introduce Physics-Informed Reward Fine-tuning (PIRF), a method that bypasses value approximation by computing trajectory-level rewards and backpropagating their gradients directly. However, a naive implementation suffers from low sample efficiency and compromised data fidelity. PIRF mitigates these issues through two key strategies: (1) a layer-wise truncated backpropagation method that leverages the spatiotemporally localized nature of physics-based rewards, and (2) a weight-based regularization scheme that improves efficiency over traditional distillation-based methods. Across five PDE benchmarks, PIRF consistently achieves superior physical enforcement under efficient sampling regimes, highlighting the potential of reward fine-tuning for advancing scientific generative modeling.

Related papers

• Scaling Reinforcement Learning for Diffusion Models via Velocity Matching [66.34541945134916]
  RewardBench fine-tuning is an important tool for adapting diffusion models to human preferences and task-specific objectives.<n>Unlike autoregressive models, diffusion models do not provide tractable likelihoods for generated samples.<n>We propose reward-based velocity matching (RVM), a simple trajectory-free update that acts directly on the velocity field.
  arXiv  Detail & Related papers  (2026-08-24T17:13:43Z)
• Decoupled Latent Optimization of Diffusion Models for Full Waveform Inversion [4.293965179144764]
  Full waveform inversion (FWI) recovers subsurface velocity from recordings by solving a severely ill-posed, nonconstrained quadratic PDE-velocity optimization.<n>Recent diffusion-prior methods improve realism at the cost of a fragile trade-off between the standard formulation and prior consistency.<n>We propose De Latent Optimization (DLO), which relaxes the latent-penalty objective into a latent-penalty objective over an auxiliary physical variable and a latent variable.<n>On the OpenFWI benchmark, DLO outperforms classical regularizers and existing diffusion-based methods under clean, noisy, and missing-trace.
  arXiv  Detail & Related papers  (2026-06-12T05:55:10Z)
• Least-Action-Guided Diffusion for Physical Extrapolation [1.2609533046091317]
  We introduce a least-action-principle-guided diffusion, LAPG, that promotes physical consistency during inference.<n>We evaluate LAPG on representative ordinary- and partial-differential-equation systems.<n>In temporal, parameter, and geometric extrapolation tests, LAPG reduces phase drift, preserves dissipative decay, captures vortex motion, and improves the lift response of airfoil flows.
  arXiv  Detail & Related papers  (2026-06-09T10:59:37Z)
• EFF-Grasp: Energy-Field Flow Matching for Physics-Aware Dexterous Grasp Generation [54.09734511705173]
  EFF-Grasp is a Flow-Matching synthesis-based framework for physics-aware dexterous grasp generation.<n>We reformulate grasp as a deterministic ordinary differential equation (ODE) process, which enables efficient and stable generation through smooth probability flows.<n> EFF-Grasp achieves superior performance in grasp quality and physical feasibility, while requiring substantially fewer sampling steps than diffusion-based baselines.
  arXiv  Detail & Related papers  (2026-03-17T06:11:56Z)
• LFPO: Likelihood-Free Policy Optimization for Masked Diffusion Models [48.68246945083386]
  Likelihood-Free Policy Optimization (LFPO) is a native framework that maps the concept of vector field flow matching to the discrete token space.<n>LFPO formulates alignment as geometric velocity rectification, which directly optimize denoising logits via contrastive updates.<n>Experiments demonstrate that LFPO not only outperforms state-of-the-art baselines on code and reasoning benchmarks but also accelerates inference by approximately 20% through reduced diffusion steps.
  arXiv  Detail & Related papers  (2026-03-02T07:42:55Z)
• Function-Space Decoupled Diffusion for Forward and Inverse Modeling in Carbon Capture and Storage [65.51149575007149]
  We present Fun-DDPS, a generative framework that combines function-space diffusion models with differentiable neural operator surrogates for both forward and inverse modeling.<n>Fun-DDPS produces physically consistent realizations free from the high-frequency artifacts observed in joint-state baselines.
  arXiv  Detail & Related papers  (2026-02-12T18:58:12Z)
• PhysE-Inv: A Physics-Encoded Inverse Modeling approach for Arctic Snow Depth Prediction [0.16206783799607727]
  We introduce PhysE-Inv, a novel framework that integrates a sophisticated sequential architecture, an LSTM-Decoder with Multi-head Attention and physics-guided contrastive learning, with physics-guided inference.<n> PhysE-Inv significantly improves prediction performance, reducing error by 20% while demonstrating superior physical consistency and resilience to data sparsity compared to empirical methods.<n>This approach pioneers a path for noise-tolerant, interpretable inverse modeling, with wide applicability in geospatial and cryospheric domains.
  arXiv  Detail & Related papers  (2026-01-23T00:43:51Z)
• Boosting Fidelity for Pre-Trained-Diffusion-Based Low-Light Image Enhancement via Condition Refinement [63.54516423266521]
  Pre-Trained Diffusion-Based (PTDB) methods often sacrifice content fidelity to attain higher perceptual realism.<n>We propose a novel optimization strategy for conditioning in pre-trained diffusion models, enhancing fidelity while preserving realism and aesthetics.<n>Our approach is plug-and-play, seamlessly integrating into existing diffusion networks to provide more effective control.
  arXiv  Detail & Related papers  (2025-10-20T02:40:06Z)
• TCPO: Thought-Centric Preference Optimization for Effective Embodied Decision-making [75.29820290660065]
  This paper proposes Thought-Centric Preference Optimization ( TCPO) for effective embodied decision-making.<n>It emphasizes the alignment of the model's intermediate reasoning process, mitigating the problem of model degradation.<n>Experiments in the ALFWorld environment demonstrate an average success rate of 26.67%, achieving a 6% improvement over RL4VLM.
  arXiv  Detail & Related papers  (2025-09-10T11:16:21Z)
• A-FloPS: Accelerating Diffusion Sampling with Adaptive Flow Path Sampler [21.134678093577193]
  A-FloPS is a principled, training-free framework for flow-based generative models.<n>We show that A-FloPS consistently outperforms state-of-the-art training-free samplers in both sample quality and efficiency.<n>With as few as $5$ function evaluations, A-FloPS achieves substantially lower FID and generates sharper, more coherent images.
  arXiv  Detail & Related papers  (2025-08-22T13:28:16Z)
• Physics-Constrained Fine-Tuning of Flow-Matching Models for Generation and Inverse Problems [3.3811247908085855]
  We present a framework for fine-tuning flow-matching generative models to enforce physical constraints and solve inverse problems in scientific systems.<n>Our approach bridges generative modelling and scientific inference, opening new avenues for simulation-augmented discovery and data-efficient modelling of physical systems.
  arXiv  Detail & Related papers  (2025-08-05T09:32:04Z)
• Iterative Distillation for Reward-Guided Fine-Tuning of Diffusion Models in Biomolecular Design [58.8094854658848]
  We address the problem of fine-tuning diffusion models for reward-guided generation in biomolecular design.<n>We propose an iterative distillation-based fine-tuning framework that enables diffusion models to optimize for arbitrary reward functions.<n>Our off-policy formulation, combined with KL divergence minimization, enhances training stability and sample efficiency compared to existing RL-based methods.
  arXiv  Detail & Related papers  (2025-07-01T05:55:28Z)
• Flow Matching Meets PDEs: A Unified Framework for Physics-Constrained Generation [21.321570407292263]
  We propose Physics-Based Flow Matching, a generative framework that embeds physical constraints, both PDE residuals and algebraic relations, into the flow matching objective.<n>We show that our approach yields up to an $8times$ more accurate physical residuals compared to FM, while clearly outperforming existing algorithms in terms of distributional accuracy.
  arXiv  Detail & Related papers  (2025-06-10T09:13:37Z)
• Provably Mitigating Overoptimization in RLHF: Your SFT Loss is Implicitly an Adversarial Regularizer [52.09480867526656]
  We identify the source of misalignment as a form of distributional shift and uncertainty in learning human preferences.<n>To mitigate overoptimization, we first propose a theoretical algorithm that chooses the best policy for an adversarially chosen reward model.<n>Using the equivalence between reward models and the corresponding optimal policy, the algorithm features a simple objective that combines a preference optimization loss and a supervised learning loss.
  arXiv  Detail & Related papers  (2024-05-26T05:38:50Z)
• Unmasking Bias in Diffusion Model Training [40.90066994983719]
  Denoising diffusion models have emerged as a dominant approach for image generation. They still suffer from slow convergence in training and color shift issues in sampling. In this paper, we identify that these obstacles can be largely attributed to bias and suboptimality inherent in the default training paradigm.
  arXiv  Detail & Related papers  (2023-10-12T16:04:41Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.