Machine Learning · Cross-listed from cs.MA
Provably Convergent Actor-Critic for MARL through Risk-aversion
Yizhou Zhang and Eric Mazumdar
Learning stationary policies in infinite-horizon general-sum Markov games (MGs) remains a fundamental open problem in Multi-
Agent Reinforcement Learning (MARL). While stationary strategies are preferred for their practicality, computing stationary forms of classic game-theoretic
equilibria is computationally intractable – a stark contrast to the comparative ease of solving single-
agent RL or zero-sum games. To bridge this gap, we study Risk-averse Quantal response
Equilibria (RQE), a solution concept rooted in behavioral game theory that incorporates risk aversion and bounded rationality. We demonstrate that RQE possesses strong regularity conditions that make it uniquely amenable to learning in MGs. We propose a novel single-timescale Actor-Critic
algorithm characterized by a faster actor and a slower critic. Leveraging the regularity of RQE, we prove that this approach achieves global convergence with finite-sample guarantees. We empirically validate our
algorithm in several environments to demonstrate superior convergence properties compared to risk-neutral baselines.
Read the full abstract ▾
Machine Learning
Why Deep Jacobian Spectra Separate: Depth-Induced Scaling and Singular-Vector Alignment
Nathanaël Haas, François Gatine, Augustin M Cosse et al. (+1 more)
Understanding why gradient-based training in deep networks exhibits strong implicit bias remains challenging, in part because tractable singular-value dynamics are typically available only for balanced deep linear models. We propose an alternative route based on two theoretically grounded and empirically testable signatures of deep Jacobians: depth-induced exponential scaling of ordered singular values and strong spectral separation. Adopting a fixed-gates view of piecewise-linear networks, where Jacobians reduce to products of masked linear maps within a single activation region, we prove the existence of Lyapunov exponents governing the top singular values at initialization, give closed-form expressions in a tractable masked model, and quantify finite-depth corrections. We further show that sufficiently strong separation forces singular-vector
alignment in
matrix products, yielding an approximately shared singular basis for intermediate Jacobians. Together, these results motivate an approximation regime in which singular-value dynamics become effectively decoupled, mirroring classical balanced deep-linear analyses without requiring balancing. Experiments in fixed-gates settings validate the predicted scaling,
alignment, and resulting dynamics, supporting a mechanistic account of emergent low-rank Jacobian structure as a driver of implicit bias.
Read the full abstract ▾
All authors Nathanaël Haas, François Gatine, Augustin M Cosse, Zied Bouraoui
Machine Learning
TFT-ACB-XML: Decision-Level Integration of Customized Temporal Fusion Transformer and Attention-BiLSTM with XGBoost Meta-Learner for BTC Price Forecasting
Raiz Ud Din and Saddam Hussain Khan
Accurate
forecasting of Bitcoin (BTC) has always been a challenge because decentralized
markets are non-linear, highly volatile, and have temporal irregularities. Existing deep learning models often struggle with interpretability and generalization across diverse
market conditions. This research presents a hybrid stacked-generalization framework, TFT-ACB-XML, for BTC closing price prediction. The framework integrates two parallel base learners: a customized Temporal Fusion
Transformer (TFT) and an
Attention-Customized Bidirectional Long Short-Term
Memory network (ACB), followed by an XGBoost regressor as the meta-learner. The customized TFT model handles long-range dependencies and global temporal dynamics via variable selection networks and interpretable single-head
attention. The ACB module uses a new
attention mechanism alongside the customized BiLSTM to capture short-term sequential dependencies. Predictions from both customized TFT and ACB are weighted through an error-reciprocal weighting strategy. These weights are derived from validation performance, where a model showing lower prediction error receives a higher weight. Finally, the framework concatenates these weighted outputs into a feature vector and feeds the vector to an XGBoost regressor, which captures non-linear residuals and produces the final BTC closing price prediction. Empirical validation using BTC data from October 1, 2014, to January 5, 2026, shows improved performance of the proposed framework compared to recent Deep Learning and
Transformer baseline models. The results show a MAPE of 0.65%, an MAE of 198.15, and an RMSE of 258.30 for one-step-ahead out-of-sample under a walk-forward evaluation on the test block. The evaluation period spans the 2024 BTC halving and the spot ETFs (exchange-traded funds) period, which coincide with major liquidity and
volatility shifts.
Read the full abstract ▾
Machine Learning
Deep Doubly Debiased Longitudinal Effect Estimation with ICE G-Computation
Wenxin Chen, Weishen Pan, Kyra Gan et al. (+1 more)
Estimating longitudinal treatment effects is essential for sequential decision-making but is challenging due to treatment-confounder feedback. While Iterative Conditional Expectation (ICE) G-computation offers a principled approach, its recursive structure suffers from error propagation, corrupting the learned outcome regression models. We propose D3-Net, a framework that mitigates error propagation in ICE training and then applies a robust final correction. First, to interrupt error propagation during learning, we train the ICE sequence using Sequential Doubly Robust (SDR) pseudo-outcomes, which provide bias-corrected targets for each regression. Second, we employ a multi-task
transformer with a
covariate simulator head for auxiliary supervision, regularizing representation learning, and a target network to stabilize training dynamics. For the final estimate, we discard the SDR correction and instead use the uncorrected nuisance models to perform Longitudinal Targeted Minimum Loss-Based Estimation (LTMLE) on the original outcomes. This second-stage, targeted debiasing ensures robustness and optimal finite-sample properties. Comprehensive experiments demonstrate that our model, D3-Net, robustly reduces bias and variance across different horizons, counterfactuals, and time-varying confoundings, compared to existing state-of-the-art ICE-based
estimators.
Read the full abstract ▾
All authors Wenxin Chen, Weishen Pan, Kyra Gan, Fei Wang
Machine Learning
Value Bonuses using Ensemble Errors for Exploration in Reinforcement Learning
Abdul Wahab, Raksha Kumaraswamy and Martha White
Optimistic value estimates provide one mechanism for directed exploration in reinforcement learning (RL). The
agent acts greedily with respect to an estimate of the value plus what can be seen as a value bonus. The value bonus can be learned by estimating a value function on reward bonuses, propagating local uncertainties around rewards. However, this approach only increases the value bonus for an action retroactively, after seeing a higher reward bonus from that state and action. Such an approach does not encourage the
agent to visit a state and action for the first time. In this work, we introduce an
algorithm for exploration called Value Bonuses with Ensemble errors (VBE), that maintains an ensemble of random action-value functions (RQFs). VBE uses the errors in the estimation of these RQFs to design value bonuses that provide first-visit optimism and deep exploration. The key idea is to design the rewards for these RQFs in such a way that the value bonus can decrease to zero. We show that VBE outperforms Bootstrap DQN and two reward bonus approaches (RND and ACB) on several classic environments used to test exploration and provide demonstrative experiments that it can scale easily to more complex environments like Atari.
Read the full abstract ▾
Machine Learning
Policy4OOD: A Knowledge-Guided World Model for Policy Intervention Simulation against the Opioid Overdose Crisis
Yijun Ma, Zehong Wang, Weixiang Sun et al. (+4 more)
The opioid epidemic remains one of the most severe public health crises in the United States, yet evaluating policy interventions before implementation is difficult: multiple policies interact within a dynamic system where targeting one risk pathway may inadvertently amplify another. We argue that effective opioid policy evaluation requires three capabilities –
forecasting future outcomes under current policies, counterfactual
reasoning about alternative past decisions, and optimization over candidate interventions – and propose to unify them through world modeling. We introduce Policy4OOD, a knowledge-guided spatio-temporal world model that addresses three core challenges: what policies prescribe, where effects manifest, and when effects unfold.Policy4OOD jointly encodes policy knowledge graphs, state-level spatial dependencies, and socioeconomic time series into a policy-conditioned
Transformer that forecasts future opioid outcomes.Once trained, the world model serves as a simulator:
forecasting requires only a forward pass, counterfactual analysis substitutes alternative policy encodings in the historical sequence, and policy optimization employs Monte Carlo Tree Search over the learned simulator. To support this framework, we construct a state-level monthly
dataset (2019–2024) integrating opioid mortality, socioeconomic indicators, and structured policy encodings. Experiments demonstrate that spatial dependencies and structured policy knowledge significantly improve
forecasting accuracy, validating each architectural component and the potential of world modeling for data-driven public health decision support. Our code and data have been released in https://github.com/antman9914/Policy4OOD.
Read the full abstract ▾
All authors Yijun Ma, Zehong Wang, Weixiang Sun, Zheyuan Zhang, Kaiwen Shi, Nitesh Chawla, Yanfang Ye
Machine Learning
A Machine Learning Approach to the Nirenberg Problem
Gianfranco Cortés, Maria Esteban-Casadevall, Yueqing Feng et al. (+4 more)
This work introduces the Nirenberg Neural Network: a numerical approach to the Nirenberg problem of prescribing Gaussian
curvature on $S^2$ for metrics that are pointwise conformal to the round metric. Our mesh-free physics-informed neural network (PINN) approach directly parametrises the conformal factor globally and is trained with a geometry-aware loss enforcing the
curvature equation. Additional consistency checks were performed via the Gauss-Bonnet
theorem, and spherical-harmonic expansions were fit to the learnt models to provide interpretability. For prescribed
curvatures with known realisability, the neural network achieves very low losses ($10^{-7} - 10^{-10}$), while unrealisable
curvatures yield significantly higher losses. This distinction enables the assessment of unknown cases, separating likely realisable functions from non-realisable ones. The current capabilities of the Nirenberg Neural Network demonstrate that neural solvers can serve as exploratory tools in
geometric analysis, offering a quantitative computational perspective on longstanding existence questions.
Read the full abstract ▾
All authors Gianfranco Cortés, Maria Esteban-Casadevall, Yueqing Feng, Jonas Henkel, Edward Hirst, Tancredi Schettini Gherardini, Alexander G. Stapleton
Machine Learning · Cross-listed from cs.GR
Variational Green's Functions for Volumetric PDEs
Joao Teixeira, Eitan Grinspun and Otman Benchekroun
Green's functions characterize the fundamental solutions of partial differential equations; they are essential for tasks ranging from shape analysis to physical simulation, yet they remain computationally prohibitive to evaluate on arbitrary
geometric discretizations. We present Variational Green's Function (VGF), a method that learns a smooth, differentiable representation of the Green's function for linear self-adjoint PDE operators, including the Poisson, the screened Poisson, and the biharmonic equations. To resolve the sharp singularities characteristic of the Green's functions, our method decomposes the Green's function into an analytic free-space component, and a learned corrector component. Our method leverages a variational foundation to impose Neumann boundary conditions naturally, and imposes Dirichlet boundary conditions via a projective layer on the output of the neural field. The resulting Green's functions are fast to evaluate, differentiable with respect to source application, and can be conditioned on other signals parameterizing our geometry.
Read the full abstract ▾
Machine Learning
Intrinsic Credit Assignment for Long Horizon Interaction
Ilze Amanda Auzina, Joschka Strüber, Sergio Hernández-Gutiérrez et al. (+3 more)
How can we train
agents to navigate uncertainty over long horizons? In this work, we propose {\Delta}Belief-RL, which leverages a
language model's own intrinsic beliefs to reward intermediate progress. Our method utilizes the change in the probability an
agent assigns to the target solution for credit assignment. By training on synthetic interaction data, {\Delta}Belief-RL teaches information-seeking capabilities that consistently outperform purely outcome-based rewards for Reinforcement Learning, with improvements generalizing to out-of-distribution applications ranging from customer service to personalization. Notably, the performance continues to improve as we scale test-time interactions beyond the training horizon, with interaction-efficiency increasing even on Pass@k metrics. Overall, our work introduces a scalable training strategy for navigating uncertainty over a long-horizon, by enabling credit assignment to intermediate actions via intrinsic {\Delta}Belief rewards.
Read the full abstract ▾
All authors Ilze Amanda Auzina, Joschka Strüber, Sergio Hernández-Gutiérrez, Shashwat Goel, Ameya Prabhu, Matthias Bethge
Machine Learning
Wireless TokenCom: RL-Based Tokenizer Agreement for Multi-User Wireless Token Communications
Farshad Zeinali, Mahdi Boloursaz Mashhadi and Rahim Tafazolli
Token Communications (TokenCom) has recently emerged as an effective new paradigm, where
tokens are the unified units of
multimodal communications and computations, enabling efficient digital semantic- and goal-oriented communications in future wireless networks. To establish a shared semantic latent space, the transmitters/receivers in TokenCom need to agree on an identical tokenizer model and codebook. To this end, an initial Tokenizer Agreement (TA) process is carried out in each communication episode, where the transmitter/receiver cooperate to choose from a set of pre-trained tokenizer models/ codebooks available to them both for efficient TokenCom. In this correspondence, we investigate TA in a multi-user downlink wireless TokenCom scenario, where the base station equipped with multiple
antennas transmits
video token streams to multiple users. We formulate the corresponding mixed-integer non-convex problem, and propose a hybrid reinforcement learning (RL) framework that integrates a deep Q-network (DQN) for joint tokenizer agreement and sub-channel assignment, with a deep deterministic policy gradient (DDPG) for beamforming. Simulation results show that the proposed framework outperforms baseline methods in terms of semantic quality and resource efficiency, while reducing the freezing events in
video transmission by 68% compared to the conventional H.265-based scheme.
Read the full abstract ▾
Machine Learning
The Appeal and Reality of Recycling LoRAs with Adaptive Merging
Haokun Liu, Gyung Hyun Je, Marco Ciccone et al. (+3 more)
The widespread availability of fine-tuned LoRA modules for open pre-trained models has led to an interest in methods that can adaptively merge LoRAs to improve performance. These methods typically include some way of selecting LoRAs from a pool and tune merging coefficients based on a task-specific
dataset. While adaptive merging methods have demonstrated improvements in some settings, no past work has attempted to recycle LoRAs found "in the wild" on model repositories like the Hugging Face Hub. To address this gap, we consider recycling from a pool of nearly 1,000 user-contributed LoRAs trained from the Llama 3.1 8B-Instruct
language model. Our empirical study includes a range of adaptive and non-adaptive merging methods in addition to a new method designed via a wide search over the methodological design space. We demonstrate that adaptive merging methods can improve performance over the base model but provide limited benefit over training a new LoRA on the same data used to set merging coefficients. We additionally find not only that the specific choice of LoRAs to merge has little importance, but that using LoRAs with randomly initialized parameter values yields similar performance. This raises the possibility that adaptive merging from recycled LoRAs primarily works via some kind of regularization effect, rather than by enabling positive cross-task transfer. To better understand why past work has proven successful, we confirm that positive transfer is indeed possible when there are highly relevant LoRAs in the pool. We release the model checkpoints and code online.
Read the full abstract ▾
All authors Haokun Liu, Gyung Hyun Je, Marco Ciccone, Zhenlin Xu, Prasanth YSS, Colin Raffel
Machine Learning
Abstractive Red-Teaming of Language Model Character
Nate Rahn, Allison Qi, Avery Griffin et al. (+3 more)
We want
language model assistants to conform to a character specification, which asserts how the model should act across diverse user interactions. While models typically follow these character specifications, they can occasionally violate them in large-scale deployments. In this work, we aim to identify types of queries that are likely to produce such character violations at deployment, using much less than deployment-level compute. To do this, we introduce abstractive red-teaming, where we search for natural-
language query categories, e.g. "The query is in Chinese. The query asks about family roles," that routinely elicit violations. These categories abstract over the many possible variants of a query which could appear in the wild. We introduce two
algorithms for efficient category search against a character-trait-specific reward model: one based on reinforcement learning on a category generator LLM, and another which leverages a strong LLM to iteratively synthesize categories from high-scoring queries. Across a 12-principle character specification and 7 target models, we find that our
algorithms consistently outperform baselines, and generate qualitatively interesting categories; for example, queries which ask Llama-3.1-8B-Instruct to predict the future lead to responses saying that AI will dominate humanity, and queries that ask GPT-4.1-Mini for essential prison survival items lead to enthusiastic recommendation of illegal weapons. Overall, we believe our results represent an important step towards realistic pre-deployment
auditing of
language model character.
Read the full abstract ▾
All authors Nate Rahn, Allison Qi, Avery Griffin, Jonathan Michala, Henry Sleight, Erik Jones
Machine Learning · Cross-listed from q-bio.QM
Free Lunch in Medical Image Foundation Model Pre-training via Randomized Synthesis and Disentanglement
Yuhan Wei, Yuting He, Linshan Wu et al. (+3 more)
Medical image foundation models (MIFMs) have demonstrated remarkable potential for a wide range of clinical tasks, yet their development is constrained by the scarcity, heterogeneity, and high cost of large-scale annotated
datasets. Here, we propose RaSD (Randomized Synthesis and Disentanglement), a scalable framework for pre-training MIFMs entirely on synthetic data. By modeling anatomical structures and appearance variations with randomized Gaussian distributions, RaSD exposes models to sufficient multi-scale structural and appearance perturbations, forcing them to rely on invariant and task-relevant anatomical cues rather than
dataset-specific textures, thereby enabling robust and transferable representation learning. We pre-trained RaSD on 1.2 million 3D volumes and 9.6 million 2D images, and extensively evaluated the resulting models across 6 imaging modalities, 48
datasets, and 56 downstream tasks. Across all evaluated downstream tasks, RaSD consistently outperforms training-from-scratch models, achieves the best performance on 17 tasks, and remains comparable to models pre-trained on large real
datasets in most others. These results demonstrate that the capacity of synthetic data alone to drive robust representation learning. Our findings establish a paradigm shift in medical AI, demonstrating that synthetic data can serve as a "free lunch" for scalable,
privacy-preserving, and clinically generalizable foundation models.
Read the full abstract ▾
All authors Yuhan Wei, Yuting He, Linshan Wu, Fuxiang Huang, Junlin Hou, Hao Chen
Machine Learning · Cross-listed from eess.IV
Visible and Hyperspectral Imaging for Quality Assessment of Milk: Property Characterisation and Identification
Massimo Martinelli, Elena Tomassi, Nafiou Arouna et al. (+8 more)
Rapid and non-destructive assessment of milk quality is crucial to ensuring both nutritional value and food safety. In this study, we investigated the potential of visible and hyperspectral imaging as cost-effective and quick-response alternatives to conventional chemical analyses for characterizing key properties of cowś milk. A total of 52 milk samples were analysed to determine their biochemical composition (polyphenols, antioxidant capacity, and fatty acids) using spectrophotometer methods and standard gas-
liquid and high-performance
liquid chromatography (GLC/HPLC). Concurrently, visible (RGB) images were captured using a standard smartphone, and hyperspectral data were acquired in the near-infrared range. A comprehensive analytical framework, including eleven different machine learning
algorithms, was employed to correlate imaging features with biochemical measurements. Analysis of visible images accurately distinguished between fresh samples and those stored for 12 days (100 percent accuracy) and achieved perfect discrimination between antibiotic-treated and untreated groups (100 percent accuracy). Moreover, image-derived features enabled perfect prediction of the polyphenols content and the antioxidant capacity using an XGBoost model. Hyperspectral imaging further achieved classification accuracies exceeding 95 percent for several individual fatty acids and 94.8 percent for treatment groups using a Random Forest model. These findings demonstrate that both visible and hyperspectral imaging, when coupled with machine learning, are powerful, non-invasive tools for the rapid assessment of milkś chemical and nutritional profiles, highlighting the strong potential of imaging-based approaches for milk quality assessment.
Read the full abstract ▾
All authors Massimo Martinelli, Elena Tomassi, Nafiou Arouna, Morena Gabriele, Laryssa Perez Fabbri, Luisa Pozzo, Bianca Castiglioni, Paola Cremonesi, Giuseppe Conte, Davide Moroni, Laura Pucci
Machine Learning
OptiML: An End-to-End Framework for Program Synthesis and CUDA Kernel Optimization
Arijit Bhattacharjee, Heng Ping, Son Vu Le et al. (+3 more)
Generating high-performance CUDA kernels remains challenging due to the need to navigate a combinatorial space of low-level transformations under noisy and expensive hardware feedback. Although large
language models can synthesize functionally correct CUDA code, achieving competitive performance requires systematic exploration and verification of optimization choices. We present OptiML, an end-to-end framework that maps either natural-
language intent or input CUDA code to performance-optimized CUDA kernels by formulating kernel optimization as search under verification. OptiML consists of two decoupled stages. When the input is natural
language, a Mixture-of-Thoughts generator (OptiML-G) acts as a proposal policy over kernel implementation strategies, producing an initial executable program. A search-based optimizer (OptiML-X) then refines either synthesized or user-provided kernels using Monte Carlo Tree Search over LLM-driven edits, guided by a hardware-aware reward derived from profiler feedback. Each candidate transformation is compiled, verified, and profiled with Nsight Compute, and evaluated by a composite objective that combines runtime with hardware bottleneck proxies and guardrails against regressions. We evaluate OptiML in both synthesis-and-optimize and optimization-only settings on a diverse suite of CUDA kernels. Results show that OptiML consistently discovers verified performance improvements over strong LLM baselines and produces interpretable optimization trajectories grounded in profiler evidence.
Read the full abstract ▾
All authors Arijit Bhattacharjee, Heng Ping, Son Vu Le, Paul Bogdan, Nesreen K. Ahmed, Ali Jannesari
Machine Learning · Cross-listed from cs.CV
UniT: Unified Multimodal Chain-of-Thought Test-time Scaling
Leon Liangyu Chen, Haoyu Ma, Zhipeng Fan et al. (+11 more)
Unified models can handle both
multimodal understanding and generation within a single architecture, yet they typically operate in a single pass without iteratively refining their outputs. Many
multimodal tasks, especially those involving complex spatial compositions, multiple interacting objects, or evolving instructions, require decomposing instructions, verifying intermediate results, and making iterative corrections. While test-time scaling (TTS) has demonstrated that allocating additional inference compute for iterative
reasoning substantially improves
language model performance, extending this paradigm to unified
multimodal models remains an open challenge. We introduce UniT, a framework for
multimodal chain-of-thought test-time scaling that enables a single unified model to reason, verify, and refine across multiple rounds. UniT combines agentic data synthesis, unified model training, and flexible test-time inference to elicit cognitive behaviors including verification, subgoal decomposition, and content
memory. Our key findings are: (1) unified models trained on short
reasoning trajectories generalize to longer inference chains at test time; (2) sequential chain-of-thought
reasoning provides a more scalable and compute-efficient TTS strategy than parallel sampling; (3) training on generation and editing trajectories improves out-of-distribution visual
reasoning. These results establish
multimodal test-time scaling as an effective paradigm for advancing both generation and understanding in unified models.
Read the full abstract ▾
All authors Leon Liangyu Chen, Haoyu Ma, Zhipeng Fan, Ziqi Huang, Animesh Sinha, Xiaoliang Dai, Jialiang Wang, Zecheng He, Jianwei Yang, Chunyuan Li, Junzhe Sun, Chu Wang, Serena Yeung-Levy, Felix Juefei-Xu
Machine Learning
Function-Space Decoupled Diffusion for Forward and Inverse Modeling in Carbon Capture and Storage
Xin Ju, Jiachen Yao, Anima Anandkumar et al. (+2 more)
Accurate characterization of subsurface flow is critical for Carbon Capture and Storage (CCS) but remains challenged by the ill-posed nature of inverse problems with sparse observations. We present Function-space Decoupled
Diffusion Posterior Sampling (Fun-DDPS), a generative framework that combines function-space
diffusion models with differentiable neural operator surrogates for both forward and inverse modeling. Our approach learns a prior distribution over geological parameters (geomodel) using a single-channel
diffusion model, then leverages a Local Neural Operator (LNO) surrogate to provide physics-consistent guidance for cross-field conditioning on the dynamics field. This decoupling allows the
diffusion prior to robustly recover missing information in parameter space, while the surrogate provides efficient gradient-based guidance for data assimilation. We demonstrate Fun-DDPS on synthetic CCS modeling
datasets, achieving two key results: (1) For forward modeling with only 25% observations, Fun-DDPS achieves 7.7% relative error compared to 86.9% for standard surrogates (an 11x improvement), proving its capability to handle extreme data sparsity where deterministic methods fail. (2) We provide the first rigorous validation of
diffusion-based inverse solvers against asymptotically exact Rejection Sampling (RS) posteriors. Both Fun-DDPS and the joint-state baseline (Fun-DPS) achieve Jensen-Shannon divergence less than 0.06 against the ground truth. Crucially, Fun-DDPS produces physically consistent realizations free from the high-frequency artifacts observed in joint-state baselines, achieving this with 4x improved sample efficiency compared to rejection sampling.
Read the full abstract ▾
All authors Xin Ju, Jiachen Yao, Anima Anandkumar, Sally M. Benson, Gege Wen
Machine Learning · Cross-listed from math.OC
Learning to Control: The iUzawa-Net for Nonsmooth Optimal Control of Linear PDEs
Yongcun Song, Xiaoming Yuan, Hangrui Yue et al. (+1 more)
We propose an optimization-informed deep neural network approach, named iUzawa-Net, aiming for the first solver that enables real-time solutions for a class of nonsmooth optimal control problems of linear partial differential equations (PDEs). The iUzawa-Net unrolls an inexact Uzawa method for saddle point problems, replacing classical preconditioners and PDE solvers with specifically designed learnable neural networks. We prove universal approximation properties and establish the asymptotic $\varepsilon$-optimality for the iUzawa-Net, and validate its promising numerical efficiency through nonsmooth
elliptic and
parabolic optimal control problems. Our techniques offer a versatile framework for designing and analyzing various optimization-informed deep learning approaches to optimal control and other PDE-constrained optimization problems. The proposed learning-to-control approach synergizes model-based optimization
algorithms and data-driven deep learning techniques, inheriting the merits of both methodologies.
Read the full abstract ▾
All authors Yongcun Song, Xiaoming Yuan, Hangrui Yue, Tianyou Zeng
Machine Learning · Cross-listed from cs.CV
MonarchRT: Efficient Attention for Real-Time Video Generation
Krish Agarwal, Zhuoming Chen, Cheng Luo et al. (+5 more)
Real-time
video generation with
Diffusion Transformers is bottlenecked by the quadratic cost of 3D self-
attention, especially in real-time regimes that are both few-step and autoregressive, where errors compound across time and each denoising step must carry substantially more information. In this setting, we find that prior sparse-
attention approximations break down, despite showing strong results for bidirectional, many-step
diffusion. Specifically, we observe that
video attention is not reliably sparse, but instead combines pronounced periodic structure driven by spatiotemporal position with dynamic, sparse semantic correspondences and dense mixing, exceeding the representational capacity of even oracle top-k
attention. Building on this insight, we propose Monarch-RT, a structured
attention parameterization for
video diffusion models that factorizes
attention using Monarch
matrices. Through appropriately aligned block structure and our extended tiled Monarch parameterization, we achieve high expressivity while preserving computational efficiency. We further overcome the overhead of parameterization through finetuning, with custom Triton kernels. We first validate the high efficacy of Monarch-RT over existing sparse baselines designed only for bidirectional models. We further observe that Monarch-RT attains up to 95%
attention sparsity with no loss in quality when applied to the state-of-the-art model Self-Forcing, making Monarch-RT a pioneering work on highly-capable sparse
attention parameterization for real-time
video generation. Our optimized implementation outperforms FlashAttention-2, FlashAttention-3, and FlashAttention-4 kernels on Nvidia RTX 5090, H100, and B200 GPUs respectively, providing kernel speedups in the range of 1.4-11.8X. This enables us, for the first time, to achieve true real-time
video generation with Self-Forcing at 16 FPS on a single RTX 5090.
Read the full abstract ▾
All authors Krish Agarwal, Zhuoming Chen, Cheng Luo, Yongqi Chen, Haizhong Zheng, Xun Huang, Atri Rudra, Beidi Chen
Machine Learning
Self-Supervised Learning via Flow-Guided Neural Operator on Time-Series Data
Duy Nguyen, Jiachen Yao, Jiayun Wang et al. (+2 more)
Self-supervised learning (SSL) is a powerful paradigm for learning from unlabeled time-series data. However, popular methods such as masked autoencoders (MAEs) rely on reconstructing inputs from a fixed, predetermined masking ratio. Instead of this static design, we propose treating the corruption level as a new degree of freedom for representation learning, enhancing flexibility and performance. To achieve this, we introduce the Flow-Guided Neural Operator (FGNO), a novel framework combining operator learning with flow matching for SSL training. FGNO learns mappings in functional spaces by using Short-Time
Fourier Transform to unify different time resolutions. We extract a rich hierarchy of features by tapping into different network layers and flow times that apply varying strengths of noise to the input data. This enables the extraction of versatile representations, from low-level patterns to high-level global features, using a single model adaptable to specific tasks. Unlike prior generative SSL methods that use noisy inputs during inference, we propose using clean inputs for representation extraction while learning representations with noise; this eliminates randomness and boosts accuracy. We evaluate FGNO across three biomedical domains, where it consistently outperforms established baselines. Our method yields up to 35% AUROC gains in neural signal decoding (BrainTreeBank), 16% RMSE reductions in skin temperature prediction (DREAMT), and over 20% improvement in accuracy and macro-F1 on SleepEDF under low-data regimes. These results highlight FGNO's robustness to data scarcity and its superior capacity to learn expressive representations for diverse time series.
Read the full abstract ▾
All authors Duy Nguyen, Jiachen Yao, Jiayun Wang, Julius Berner, Animashree Anandkumar
Machine Learning · Cross-listed from cs.CL
Tunyu Zhang, Xinxi Zhang, Ligong Han et al. (+10 more)
Diffusion large
language models (DLLMs) have emerged as powerful generative models with the promise of fast text generation through parallel decoding. However, realizing this potential in practice remains challenging: reducing the number of decoding steps, typically causes a substantial degradation in output quality due to
token factorization error. To alleviate this, we propose a self-
distillation framework that trains a few-step student to match the generative trajectory of a full-step teacher. We theoretically and empirically show that trajectory-level supervision mitigates this factorization error, thereby enabling effective few-step decoding. We further incorporate Direct Discriminative Optimization (DDO), a reverse-KL objective that encourages mode-seeking toward the teacher's modes, yielding stronger performance on challenging
reasoning tasks. Across
reasoning and code-generation
benchmarks, our method substantially narrows the gap between few-step and full-step decoding. The source code is available at https://github.com/Tyrion58/T3D.
Read the full abstract ▾
All authors Tunyu Zhang, Xinxi Zhang, Ligong Han, Haizhou Shi, Xiaoxiao He, Zhuowei Li, Hao Wang, Kai Xu, Akash Srivastava, Chengzhi Mao, Hao Wang, Vladimir Pavlovic, Dimitris N. Metaxas
Machine Learning · Cross-listed from cs.AI
Think like a Scientist: Physics-guided LLM Agent for Equation Discovery
Jianke Yang, Ohm Venkatachalam, Mohammad Kianezhad et al. (+2 more)
Explaining observed phenomena through symbolic, interpretable formulas is a fundamental goal of science. Recently, large
language models (LLMs) have emerged as promising tools for symbolic equation discovery, owing to their broad domain knowledge and strong
reasoning capabilities. However, most existing LLM-based systems try to guess equations directly from data, without modeling the multi-step
reasoning process that scientists often follow: first inferring physical properties such as
symmetries, then using these as priors to restrict the space of candidate equations. We introduce KeplerAgent, an agentic framework that explicitly follows this scientific
reasoning process. The
agent coordinates physics-based tools to extract intermediate structure and uses these results to configure symbolic regression engines such as PySINDy and PySR, including their function libraries and structural constraints. Across a suite of physical equation
benchmarks, KeplerAgent achieves substantially higher symbolic accuracy and greater robustness to noisy data than both LLM and traditional baselines.
Read the full abstract ▾
All authors Jianke Yang, Ohm Venkatachalam, Mohammad Kianezhad, Sharvaree Vadgama, Rose Yu
Machine Learning · Cross-listed from cs.GT
Is Online Linear Optimization Sufficient for Strategic Robustness?
Yang Cai, Haipeng Luo, Chen-Yu Wei et al. (+1 more)
We consider bidding in repeated
Bayesian first-price auctions. Bidding
algorithms that achieve optimal regret have been extensively studied, but their strategic robustness to the seller's manipulation remains relatively underexplored. Bidding
algorithms based on no-swap-regret
algorithms achieve both desirable properties, but are suboptimal in terms of statistical and computational efficiency. In contrast, online gradient ascent is the only
algorithm that achieves $O(\sqrt{TK})$ regret and strategic robustness [KSS24], where $T$ denotes the number of auctions and $K$ the number of bids. In this paper, we explore whether simple online linear optimization (OLO)
algorithms suffice for bidding
algorithms with both desirable properties. Our main result shows that sublinear linearized regret is sufficient for strategic robustness. Specifically, we construct simple black-box reductions that convert any OLO
algorithm into a strategically robust no-regret bidding
algorithm, in both known and unknown value distribution settings. For the known value distribution case, our reduction yields a bidding
algorithm that achieves $O(\sqrt{T \log K})$ regret and strategic robustness (with exponential improvement on the $K$-dependence compared to [KSS24]). Moreover, we can further exploit the single-strategy
benchmark and construct a bidding
algorithm with $O(\sqrt{T(\log K+\log T)})$ regret and $O(1)$ strategic robustness. For the unknown value distribution case, our reduction gives a bidding
algorithm with high-probability $O(\sqrt{T (\log K+\log(T/\delta)})$ regret and strategic robustness, while removing the bounded density assumption made in [KSS24].
Read the full abstract ▾
All authors Yang Cai, Haipeng Luo, Chen-Yu Wei, Weiqiang Zheng
Machine Learning
Community Concealment from Graph Neural Networks
Dalyapraz Manatova, Pablo Moriano and L. Jean Camp
Graph neural networks (GNNs) enable powerful unsupervised learning of communities. However, such inference may inadvertently expose sensitive group structures, critical clustered patterns, or collective behaviors, raising concerns about sensitive group-level
privacy. In social and critical infrastructure networks, unauthorized community inference can reveal coordinated asset groups, operational hierarchies, and system dependencies that may be exploited for reconnaissance or profiling. We study a defensive setting in which a network (or defender) operator seeks to conceal a community of interest while making only small, utility-preserving modifications to the network. Our analysis shows that community concealment depends on two measurable factors: the connectivity at the community boundary and the feature similarity between the protected community and its neighbors. Guided by these observations, we introduce Feature-Community-guided DICE (FCom-DICE), a perturbation strategy built on DICE (Disconnect Internally Connect Externally) that rewires a set of structurally influential edges and adjusts node features to reduce the distinctiveness exploited by GNN message passing. Across synthetic
benchmarks and real network graphs such as Facebook, Wikipedia, and Bitcoin Transactions, FCom-DICE consistently outperforms structure-only DICE under the same perturbation budgets. The largest improvements are observed for communities that are weakly connected to the rest of the network and well separated in feature space. These gains are achieved while preserving key structural and feature characteristics of the original network. These results demonstrate the effectiveness of feature-aware perturbations for reducing the recoverability of targeted communities under GNN-based community inference.
Read the full abstract ▾
Machine Learning
ExtractBench: A Benchmark and Evaluation Methodology for Complex Structured Extraction
Nick Ferguson, Josh Pennington, Narek Beghian et al. (+4 more)
Unstructured documents like PDFs contain valuable structured information, but downstream systems require this data in reliable, standardized formats. LLMs are increasingly deployed to automate this extraction, making accuracy and reliability paramount. However, progress is bottlenecked by two gaps. First, no end-to-end
benchmark evaluates PDF-to-JSON extraction under enterprise-scale schema breadth. Second, no principled methodology captures the semantics of nested extraction, where fields demand different notions of correctness (exact match for identifiers, tolerance for quantities, semantic equivalence for names), arrays require
alignment, and omission must be distinguished from
hallucination. We address both gaps with ExtractBench, an open-source
benchmark and evaluation framework for PDF-to-JSON structured extraction. The
benchmark pairs 35 PDF documents with JSON Schemas and human-annotated gold labels across economically valuable domains, yielding 12,867 evaluatable fields spanning schema complexities from tens to hundreds of fields. The evaluation framework treats the schema as an executable specification: each field declares its scoring metric. Baseline evaluations reveal that frontier models (GPT-5/5.2, Gemini-3 Flash/Pro, Claude 4.5 Opus/Sonnet) remain unreliable on realistic schemas. Performance degrades sharply with schema breadth, culminating in 0% valid output on a 369-field financial reporting schema across all tested models. We release ExtractBench at https://github.com/ContextualAI/extract-bench.
Read the full abstract ▾
All authors Nick Ferguson, Josh Pennington, Narek Beghian, Aravind Mohan, Douwe Kiela, Sheshansh Agrawal, Thien Hang Nguyen
Machine Learning
Intrinsic-Energy Joint Embedding Predictive Architectures Induce Quasimetric Spaces
Anthony Kobanda and Waris Radji
Joint-
Embedding Predictive Architectures (JEPAs) aim to learn representations by predicting target
embeddings from context
embeddings, inducing a scalar compatibility energy in a latent space. In contrast, Quasimetric Reinforcement Learning (QRL) studies goal-conditioned control through directed distance values (cost-to-go) that support reaching goals under asymmetric dynamics. In this short article, we connect these viewpoints by restricting
attention to a principled class of JEPA energy functions : intrinsic (least-action) energies, defined as infima of accumulated local effort over admissible trajectories between two states. Under mild closure and additivity assumptions, any intrinsic energy is a quasimetric. In goal-reaching control, optimal cost-to-go functions admit exactly this intrinsic form ; inversely, JEPAs trained to model intrinsic energies lie in the quasimetric value class targeted by QRL. Moreover, we observe why symmetric finite energies are structurally mismatched with one-way reachability, motivating asymmetric (quasimetric) energies when directionality matters.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CL
Moonshine v2: Ergodic Streaming Encoder ASR for Latency-Critical Speech Applications
Manjunath Kudlur, Evan King, James Wang et al. (+1 more)
Latency-critical
speech applications (e.g., live transcription, voice commands, and real-time translation) demand low time-to-first-
token (TTFT) and high transcription accuracy, particularly on resource-constrained edge devices. Full-
attention Transformer encoders remain a strong accuracy baseline for automatic
speech recognition (ASR) because every frame can directly attend to every other frame, which resolves otherwise locally ambiguous acoustics using distant lexical context. However, this global dependency incurs quadratic complexity in sequence length, inducing an inherent "encode-the-whole-utterance" latency profile. For streaming use cases, this causes TTFT to grow linearly with utterance length as the encoder must process the entire prefix before any decoder
token can be emitted. To better meet the needs of on-device, streaming ASR use cases we introduce Moonshine v2, an ergodic streaming-encoder ASR model that employs sliding-window self-
attention to achieve bounded, low-latency inference while preserving strong local context. Our models achieve state of the art word error rates across standard
benchmarks, attaining accuracy on-par with models 6x their size while running significantly faster. These results demonstrate that carefully designed local
attention is competitive with the accuracy of full
attention at a fraction of the size and latency cost, opening new possibilities for interactive
speech interfaces on edge devices.
Read the full abstract ▾
All authors Manjunath Kudlur, Evan King, James Wang, Pete Warden
Machine Learning
Olmix: A Framework for Data Mixing Throughout LM Development
Mayee F. Chen, Tyler Murray, David Heineman et al. (+5 more)
Data mixing – determining the ratios of data from different domains – is a first-order concern for training
language models (LMs). While existing mixing methods show promise, they fall short when applied during real-world LM development. We present Olmix, a framework that addresses two such challenges. First, the configuration space for developing a mixing method is not well understood – design choices across existing methods lack justification or consensus and overlook practical issues like data constraints. We conduct a comprehensive empirical study of this space, identifying which design choices lead to a strong mixing method. Second, in practice, the domain set evolves throughout LM development as
datasets are added, removed, partitioned, and revised – a problem setting largely unaddressed by existing works, which assume fixed domains. We study how to efficiently recompute the mixture after the domain set is updated, leveraging information from past mixtures. We introduce mixture reuse, a mechanism that reuses existing ratios and recomputes ratios only for domains affected by the update. Over a sequence of five domain-set updates mirroring real-world LM development, mixture reuse matches the performance of fully recomputing the mix after each update with 74% less compute and improves over training without mixing by 11.6% on downstream tasks.
Read the full abstract ▾
All authors Mayee F. Chen, Tyler Murray, David Heineman, Matt Jordan, Hannaneh Hajishirzi, Christopher Ré, Luca Soldaini, Kyle Lo
Machine Learning
Categorical Flow Maps
Daan Roos, Oscar Davis, Floor Eijkelboom et al. (+5 more)
We introduce Categorical Flow Maps, a flow-matching method for accelerated few-step generation of categorical data via self-
distillation. Building on recent variational formulations of flow matching and the broader trend towards accelerated inference in
diffusion and flow-based models, we define a flow map towards the simplex that transports probability mass toward a predicted endpoint, yielding a parametrisation that naturally constrains model predictions. Since our trajectories are continuous rather than discrete, Categorical Flow Maps can be trained with existing
distillation techniques, as well as a new objective based on endpoint consistency. This continuous formulation also automatically unlocks test-time inference: we can directly reuse existing guidance and reweighting techniques in the categorical setting to steer sampling toward downstream objectives. Empirically, we achieve state-of-the-art few-step results on images, molecular graphs, and text, with strong performance even in single-step generation.
Read the full abstract ▾
All authors Daan Roos, Oscar Davis, Floor Eijkelboom, Michael Bronstein, Max Welling, İsmail İlkan Ceylan, Luca Ambrogioni, Jan-Willem van de Meent
Machine Learning
Diffusion Alignment Beyond KL: Variance Minimisation as Effective Policy Optimiser
Zijing Ou, Jacob Si, Junyi Zhu et al. (+4 more)
Diffusion alignment adapts pretrained
diffusion models to sample from reward-tilted distributions along the denoising trajectory. This process naturally admits a Sequential Monte Carlo (SMC) interpretation, where the denoising model acts as a proposal and reward guidance induces importance weights. Motivated by this view, we introduce Variance Minimisation Policy Optimisation (VMPO), which formulates
diffusion alignment as minimising the variance of log importance weights rather than directly optimising a Kullback-Leibler (KL) based objective. We prove that the variance objective is minimised by the reward-tilted target distribution and that, under on-policy sampling, its gradient coincides with that of standard KL-based
alignment. This perspective offers a common lens for understanding
diffusion alignment. Under different choices of potential functions and variance minimisation strategies, VMPO recovers various existing methods, while also suggesting new design directions beyond KL.
Read the full abstract ▾
All authors Zijing Ou, Jacob Si, Junyi Zhu, Ondrej Bohdal, Mete Ozay, Taha Ceritli, Yingzhen Li