Reservoir Engineering and Simulation Methods

Neueste Paper

• Cool the Sampler, Not the Learner: Sampling Temperature Moves the Staleness Cliff of Importance-Corrected GRPO

  Taiheng Pan · 1. Oktober 2026

  Production RL for language models lets the sampler fall behind the learner and repairs the resulting mismatch with a truncated importance weight. We ask how long the sampler can go without a refresh under that correction, and find a cliff: on Qwen2.5-Math-1.5B and GSM8K, importance-corrected GRPO re…

• AneumoBench: A Source-Linked Benchmark for Synthetic-Geometry Transfer in Aneurysm CFD

  Xigui Li, Yuanye Zhou, Feiyang Xiao, Xin Guo, Chen Jiang, Tan Pan, Xingmeng Zhang, Cenyu Liu, Zeyun Miao, Xiansheng Wang, Qimeng Wang, Yichi Zhang, Wenbo Zhang, Hongwei Zhang, Ruoxi Jiang, Fengping Zhu, Limei Han, Chensen Lin, Yuan Cheng · 1. Oktober 2026

  Scientific machine learning uses simulation data to train surrogate models for fast physical-field prediction across geometries. Local shape editing can expand limited geometry collections, but whether its variants improve prediction on unseen geometries, and how to allocate them across sources, req…

• Fenchel Tilting: Weighted Correction for Efficient Finetuning of Generative Models

  Maksim Bobrin, Maksim Zhdanov, Dmitry Dylov · 1. Oktober 2026

  Adapting a pretrained generative model to an arbitrary preference expressed as a utility function underlies reward alignment, guided design, and constraint satisfaction, enabling diverse applications. Existing fine-tuning methods trade off generality against computational cost: they either restrict …

• GFD-OPD: Guidance-Folded On-Policy Distillation of Diffusion Models Across Scales

  Zhenxing Zhang, Jiayan Teng, Wenxu Wu, Zhuoyi Yang, Jiazheng Xu, Wendi Zheng, Jie Tang, Dan Guo, Meng Wang · 1. Oktober 2026

  On-policy distillation (OPD) has demonstrated two important capabilities in language models: compressing large teachers into smaller students and merging expert models into a single model. Existing diffusion OPD, however, mostly focus on the latter, with teachers and students sharing the same backbo…

• Towards Universal Wasserstein Barycenters through Flow Matching

  Eduardo Fernandes Montesuma · 1. Oktober 2026

  Defining a weighted mean over probability measures under probability metrics is a central tool in probabilistic machine learning. Under the Wasserstein metric, these are called \emph{Wasserstein barycenters}. While most approaches compute barycenters for a fixed weight vector, approximating the whol…

• MeanFlowAdvantage: Stable Reward Fine-Tuning for Few-Step Average-Velocity Generators

  Haocheng Tang, Tianchi Xie, Xingqiao Lin · 30. September 2026

  MeanFlow enables efficient few-step generation by predicting interval-average velocities, but this representation creates a mismatch for reward fine-tuning: existing advantage-based objectives are typically defined on instantaneous velocities or equivalent $x_0$-space predictions, whereas inference …

• Do We Really Need KL Divergence for On-Policy Distillation of Large Language Models?

  Wenze Lin, Jiyuan Long, Jiale Zhao, Shenzhi Wang, Xitai Jiang, Ce Luo, Rui Lan, Qianli Ma, Fukang Wen, Hui Wu, Liyuan Chen, Shuoling Liu, Jiangpeng Yan, Gao Huang · 29. September 2026

  Since the advent of knowledge distillation, KL divergence has been the standard loss in distillation. Recently, on-policy distillation (OPD) has emerged as an efficient post-training paradigm for LLMs. As a distillation method, OPD naturally inherits KL divergence as its standard loss. However, in t…

• Activation Flow: Manufacturing Activations for Steering

  Hong Kiat Tan, Linh Le, David Williams-King · 29. September 2026

  Difference-in-means steering requires activations recorded while a model shows the desired behavior, which a sandbagging model withholds by deliberately underperforming. We introduce Activation Flow (ActFlow), which manufactures these activations from $k$ correct labels without fine-tuning. ActFlow …

• A Flow Matching Framework for Neural Representational Dissimilarity

  Zeyuan Ye, Xue-Xin Wei · 28. September 2026

  Neural representational dissimilarity quantifies differences between neural response distributions, and is essential for comparing neural codes across stimuli, brain areas, tasks, and models. Commonly used distance metrics involve different assumptions and are estimated with separate methods. Here, …

• WTF?! Simulation-Free Reinforcement Learning with Wasserstein-Tilted Flow Maps

  Abbas Mammadov, Jerry Y. Huang, Justin Lin, Partha Kaushik, Sheel Shah, Kartik Nair, Yee Whye Teh, Nicholas M. Boffi · 24. September 2026

  Reward fine-tuning aims to update a pre-trained flow-based generative model to improve the downstream reward of its generated samples. Existing methods typically formulate this problem as sampling from a reward-tilted distribution, the solution to a KL-regularized reward-maximization problem. Here, …

• Can You Delete a Year of Market Data? Machine Unlearning Against Exact Retraining Oracles

  Junyi Ye · 23. September 2026

  When a data license expires, deleting stored records does not remove influence encoded in a trained forecaster. Machine unlearning seeks to remove this influence without retraining. We benchmark temporal unlearning with 3,200 paired references trained on all data and oracles retrained without the re…

• Entropy Can Flow, or It Can Guide. Be Entropy. LEDFlow: Introducing Entropy-guided Generation Order into Uniform Discrete Flow

  Tung Sum Thomas Kwok, Yidong Ouyang, Yingjia Wan, Ying Nian Wu, Zhijiang Guo, Oscar Leong · 23. September 2026

  Uniform discrete flow permits repeated updates at every generation position. While continued revision supports correction of wrong tokens, it also exposes correct intermediate predictions to later errors. An experiment on Sudoku puzzles shows that 9.4% of generated cells are correct at an intermedia…

• Terminal Shrinkage Averaging Reveals a Schedule-Estimator Interaction in LLM Pretraining

  Adam Ousherovitch, Yixin Wang · 23. September 2026

  Large language model (LLM) pretraining conventionally returns the raw final iterate. This couples two design choices: the learning-rate schedule that generates the parameter trajectory and the estimator that constructs the deployed model (e.g. the raw final iterate or a checkpoint average). A schedu…

• When Is Availability-Aware Training Worth It? A Benchmark and Empirical Study of Interruption-Resilient Optimization Under Predictable Compute Schedules

  Subhadip Mitra · 22. September 2026

  Training under non-stationary but predictable compute availability (satellites under eclipse, duty-cycled edge devices, power-capped datacenters) is often framed as needing specialized, availability-aware optimizers. We test that premise. We release OrbitTrace, a benchmark of 50 physics-grounded ava…

• Whitening Inverts the Hierarchy: What the Norm of a Whitened Embedding Measures

  Mohammed Ahnouch, Lotfi Elaachak · 22. September 2026

  Whitening a foundation-model embedding and using its squared norm as a training-free likelihood surrogate is motivated by the observation that whitened coordinates often appear approximately standard normal. We show that this observation follows from the projection central limit theorem and therefor…

• Classifier-Free Guidance in Flow Matching: Non-Autonomous Potentials, Overshoot, and Posterior-Mean Control

  Jishen Peng, Zheng Ma · 22. September 2026

  Classifier-free guidance (CFG) improves conditional generation in Flow Matching, but strong guidance can distort the generated distribution and reduce diversity. We provide a geometric account of this behavior by viewing Flow Matching as a time-varying gradient flow and characterizing how CFG reshap…

• $\lambda$-Controlled GRPO: Turning Flow-Matching Ratio Instability into a Budgeted Resource

  Yufeng Wang, Parivesh Priye, Meeshawn Marathe, Ramit Pahwa · 21. September 2026

  Reinforcement learning is increasingly used to align image generators with reward signals, and Flow-GRPO recently extended this paradigm to flow-matching models by treating the denoising sampler as a stochastic policy that can be optimized from reward feedback. Training in this setting is unstable i…

• A Generative-AI Modeling Framework for Explainable Decision Support in Complex Geosteering Scenarios

  Sergey Alyaev, Kristian Fossum, Hibat Errahmen Djecta, Jan Tveranger, Ahmed H. Elsheikh · 18. September 2026

  The real-time process of directional changes while drilling, known as geosteering, is crucial for hydrocarbon extraction and emerging directional drilling applications such as geothermal energy, civil infrastructure, and CO2 storage. The geo-energy industry seeks an automatic geosteering workflow th…

• FlowSGS: Improving Flow Matching Priors for Inverse Imaging with Stochastic Interpolants

  Tianao Li, Xinhui Qian, Emma Alexander · 18. September 2026

  Flow matching has emerged as the state-of-the-art generative model and has been used for plug-and-play (PnP) priors to solve inverse problems in computational imaging. However, existing flow-based inverse solvers assume linear forward models and/or make simplifying approximations in posterior sampli…

• How to Guide Your Language Flow

  Rohit Dilip, Tianrong Chen, Yuyang Wang, David Van Valen, Joshua Susskind, Miguel Angel Bautista · 18. September 2026

  We introduce a new method to guide flow matching models. Our approach, which we call probe guidance, uses the frozen internal states of an existing diffusion model to construct a guidance signal. This works using a similar principle as autoguidance, but eliminates the need for an additional forward …

• DART: Distillation-Aware Reparameterization for Training-Free LoRA Reuse in Few-Step Video Diffusion Models

  Shihong Li, Juntao Xu, JinCao, Maowen Tang, Jun Huang, Jintao Li · 18. September 2026

  Step distillation reduces the cost of video generation, but reusing a LoRA trained for a longer trajectory can alter its functional effect or degrade target quality. Static parameter compatibility offers one perspective on this problem; our observations show that similar measured geometry can coexis…

• Right Direction, Wrong Step: Geometric Analysis of Finite-Step Failure in Looped Transformers

  Zhihao Guo, Zonghan Wu, Haizhou Du, Huan Huo, Yilei Shao, Athanasios V. Vasilakos, Qingsong Wen · 16. September 2026

  Looped Transformers offer a parameter-efficient route to test-time scaling by reusing shared layers for iterative latent reasoning. However, additional iterations can reduce support for a reference answer, leaving unclear whether an update's direction is locally unhelpful or its full displacement mo…

• How I learned to stop worrying and love StopGrads: Stationarity, Convergence, and a case study on Flow Map Learning

  Max W. Shen, Mark Goldstein, Zichu Wang, Aahlad Puli, Rajesh Ranganath · 16. September 2026

  Stopgrads are widely used in training machine learning models, but stopgrads can alter the gradient, stationary points and convergence guarantees of the original objective, which can make stopgrad training theoretically ungrounded. We introduce a stopgrad regression principle, which identifies a gen…

• Branched Optimal Transport Amortization

  Semyon Semenov, Viktor Kovalchuk, Meir Roketlishvili, Albert Baichorov, Fakhri Karray, Martin Takac, Arip Asadulaev · 15. September 2026

  Methods of Branched Optimal Transport (BOT) mimic the economy and efficiency of natural tree-like structures, such as those found in rivers and biological systems. These methods are widely applicable for designing efficient networks in society, from river basins and blood vessels to mail and gas dis…

• SCOPE-OPSD: Fisher-Conditioned Privileged Subspaces for On-Policy Self-Distillation

  Yunmeng Chen (Chongqing Ant Consumer Finance Co., Ltd), Kunyu Wang (Alibaba Cloud Computing Co., Ltd), Peihan Li (Chongqing Ant Consumer Finance Co., Ltd), Yi Wang (Chongqing Ant Consumer Finance Co., Ltd), Shuyin Xia (Chongqing University of Posts and Telecommunications), Yi Liu (Chongqing Ant Consumer Finance Co., Ltd), Xinyong Cheng (Alibaba Cloud Computing Co., Ltd), Dehui Wang (Alibaba Cloud Computing Co., Ltd), Xiangyong Zhai (Alibaba Cloud Computing Co., Ltd), Yanxing Liu (Chongqing Ant Consumer Finance Co., Ltd), Song Liu (Chongqing Ant Consumer Finance Co., Ltd) · 14. September 2026

  On-policy self-distillation (OPSD) scores student-generated prefixes with a solution-conditioned self-teacher, yet transfers supervision only through next-token probabilities. We ask whether the aligned final-layer discrepancy offers a useful second channel, and how to test that channel without conf…