arXiv subjects

Maziar Raissi

Publications and source records attributed to Maziar Raissi.

At least 19 recordsLinked to original sources

arXiv2026

Beyond Compilation: Evaluating Faithful Natural-Language-to-Lean Statement Formalization

Lean verifies that a generated declaration is well typed, but not that it states what the user intended. For statement autoformalization without canonical Lean targets, we study two questions: how far an LLM-based semantic criterion can be trusted, and how much compilation overstates faithfulness across systems. Our criterion requires Lean compilation and agreement of GPT-5.2 and Gemini-2.5-Pro. On an independently audited random sample, it agrees with the human majority on 91.5\% of cases (Wilson 95\% CI: 81.6--96.3\%); humans confirm 95.7\% of the outputs it accepts and 77--81\% of the outputs it rejects, so the criterion is reliable in aggregate and errs on the conservative side. A comparison with LeanScorer, an independent third-family judge, and a BEq formal cross-check support the same picture. Across eight systems evaluated on 227 graduate-level statements, the compile--faithfulness gap varies widely with the system, from 1.3 percentage points for one-shot Gemini-2.5-Pro to 29.5 points for a tool-augmented GPT-5.2 agent, which compiles 87.2\% of statements but is faithful on 57.7\%. A $2^3$ tool ablation of this agent shows that Lean feedback drives most of its gain in compilation, and nearly half of that gain consists of outputs that fail the semantic criterion.

cs.AI↗

arXiv2026

Online Optimization with Unknown Time-Varying Parameters Using Noisy Gradient Measurements

We study online optimization problems in which the cost function depends on latent, time-varying parameters that are unmeasurable and governed by unknown dynamics. Specifically, we consider a strongly convex cost function whose linear term evolves according to unknown linear stochastic dynamics about an unknown nominal operating point, while the algorithm has access only to finite noisy gradient measurements. We propose a solution that uses control theoretic tools to reconstruct the latent parameters from gradient observations using a Gauss-Markov estimator, then identifies the parameter dynamics using a bias-reduced instrumental variable estimator, and finally forecasts the parameters to compute the minimizer. We provide a high probability, finite-horizon bound on the expected tracking error. We illustrate our algorithm on a series of numerical examples.

math.OC↗

arXiv2026

ToolGate: An Executable Acceptance Pipeline for Tool-Dependent Scientific Benchmark Construction

Scientific benchmarks are commonly built by domain experts who write tasks and cross-check one another's work, or who adapt existing material from textbooks, published papers, and online resources. These routes can produce strong evaluations, but they require substantial per-item labor. Language models can reduce this repeated work by proposing candidates quickly. The remaining problem is acceptance. We target scientific questions whose answers require computations with specialist software rather than unaided reasoning alone. A candidate is invalid if its script fails or returns a different answer, or trivial if a model answers it without the software. We present ToolGate, which treats every generated item as a proposal and keeps it only if three gates pass. First, an executable solution script must reproduce the proposed answer when run with the scientific software. Second, randomized no-tool screening rejects candidates that models can already solve from the prompt alone. Third, a tool-using agent must solve each survivor within a fixed time limit. We instantiate ToolGate in FEniCSx with 500 generation attempts. The local-verification gate retains 478 candidates. For final reporting, we rescreen this pool after generation: two randomized no-tool screens exclude 222 from the reported pool, and direct GPT-5.5 API calls at medium reasoning (the API default) exclude another 121. Of the remaining 135, a GPT-5.5 Codex CLI agent with access to FEniCSx solves 130; exact deduplication leaves 128 unique protocol survivors. ToolGate turns repeated answer checking and difficulty screening into an auditable process while leaving domain design and final review to experts.

cs.AI↗

arXiv2026

From Centerlines to Hemodynamics: Anisotropic RBF Decoders for Coronary Arteries

Accurate and rapid estimation of hemodynamic metrics, such as pressure and wall shear stress (WSS), is important for assessing the severity of Coronary Artery Disease (CAD). Existing approaches, including invasive Fractional Flow Reserve (FFR) measurements and computationally expensive Computational Fluid Dynamics (CFD) simulations, face challenges in invasiveness, cost, and speed. We present a learned surrogate for fast prediction of CFD-simulated coronary hemodynamics from vessel centerline geometry. The model encodes 1D vessel centerlines together with inlet flow rate using a transformer-based encoder, and predicts continuous wall-based fields via an anisotropic Radial Basis Function (RBF) decoder aligned with vessel morphology. To support training and evaluation, we introduce two datasets with paired steady-state OpenFOAM simulations: (i) a synthetic benchmark of $4{,}200$ single-vessel geometries with controlled anatomical variations, and (ii) a multi-vessel dataset derived from ImageCAS including $4{,}800$ cases spanning both right and left coronary arteries, generated by randomly introducing stenoses and varying physiologically plausible flow rates. Across both datasets, our method achieves lower pressure and WSS errors than strong neural-operator baselines (GNOT, Transolver, and ONO) at a fraction of the computational cost of CFD. On the multi-vessel dataset, using $1{,}024$ anisotropic RBF centers our model reduces the mean relative $\ell_2$ error by $52\%$ compared to the best neural-operator baseline, while at $128$ centers it requires $13.8\times$ fewer FLOPs than GNOT and still outperforms all neural-operator baselines. The single-vessel dataset is publicly available at https://huggingface.co/datasets/angioinsight/single-vessel-flow

cs.CE↗

arXiv2026

Backtrader-Bench: Benchmarking LLM Agents on Algorithmic Trading with Self-Generated MCQs

Evaluating LLM coding agents in algorithmic trading is difficult because static benchmarks risk data contamination and numerical backtest outputs require ground truth from actual code execution. We present Backtrader-Bench, a framework with two complementary pipelines. A deterministic multiple-choice question (MCQ) pipeline generates questions from backtest configurations across five trading strategies, 33 templates, and three difficulty tiers, with an independent checker that re-derives every answer. A generator-solver filtering pipeline autonomously mines harder questions: a generator writes questions verified by executable code, converts them to MCQs, and discards any that a no-tool solver can answer without code execution. We evaluate 11 models without tools (10 runs each) and four with-tools configurations on a 30-question curated set. Tool-augmented agents reach 90.0% accuracy in a single pass (GPT-5.5 and Opus 4.7), outperforming the best no-tools baselines (73.0%, averaged over 10 runs) by 17 percentage points. On 38 separately mined questions, no-tools accuracy drops further, with half the models falling to roughly random-chance level (25%). Beyond evaluation, the scalable MCQ infrastructure is designed to produce a training corpus for reinforcement learning, with the ultimate goal of building a specialized agent for quantitative trading workflows.

cs.CL↗

arXiv2026

VeraGrid-Agent: Tool-Augmented LLMs for Distribution Optimal Power Flow at the Grid Edge

Language models have demonstrated remarkable success in solving a wide range of tasks. However, answering complex scientific questions about the power flow often requires solving the distribution optimal power flow (D-OPF) problem. These questions call for numerical solvers and simulators, as linguistic reasoning from parametric knowledge often gives incorrect answers. In this work, we present VeraGrid-Agent, a tool-augmented LLM that autonomously writes the simulator input, executes the open-source VeraGrid solver, and reads the solver output before answering. To evaluate performance, we introduce VeraGrid-MCQ-150, a set of deterministic, expert template driven, $150$ multiple-choice questions. We evaluate the performance under two regimes: (i) no-tool reasoning and (ii) agent (LLM with simulator access). Without tools, every model performs with an accuracy of $42.7\%$--$49.3\%$. However, with VeraGrid-Agent, accuracy increases to $97.3\%$--$100.0\%$. We also do a failure-mode analysis to show that the few remaining errors arise from wrong interpretations during multi-step reasoning, rather than any failure in the simulators execution.

eess.SY↗

arXiv2026

Spectrogram-Based Joint Detection, Localization, and Classification of Events in Continuously Recorded IBR Waveforms

Continuously recorded high-resolution waveform measurements provide rich information about fast power system dynamics. However, they require automated methods to identify events. This problem is addressed by developing a spectrogram-based framework to jointly detect, localize, and classify events in real-world continuously recorded waveforms at the terminal of an Inverter-Based Resource. We recast this problem as a temporal object detection problem on spectrogram images, as they capture the transient and harmonic signatures more explicitly than in raw waveform data. Each time-series waveform is transformed using the short-time Fourier transform, and the resulting per-channel spectrograms are stacked as a tensor for event detection. We benchmark this method against a detector operating directly on raw time-series measurements. Experiments on single-phase disturbances and three-phase faults demonstrate that the proposed spectrogram method consistently improves event detection, localization, and classification over the raw waveform baseline.

cs.SD↗

arXiv2026

PHREEQC-MCQ-200: A Diagnostic Benchmark for Tool-Augmented Scientific Simulator Agents

Large language model agents are increasingly connected to scientific software, yet it remains unclear when tool access makes scientific computation more reliable rather than merely more complex. We introduce PHREEQC-MCQ-200, a benchmark for evaluating tool-augmented agents on deterministic aqueous-geochemistry simulations. The benchmark contains 200 multiple-choice questions derived from 21 validated PHREEQC scenarios, requiring agents to construct simulator inputs, execute PHREEQC, inspect structured outputs, and commit to final answers. Across multiple frontier and mid-tier model families, simulator access substantially improves aggregate accuracy, confirming that grounded execution is necessary for many scientific-computation tasks. However, the gains are not monotonic: tool-augmented agents also lose items they answered correctly without tools, revealing regressions that average accuracy alone hides. We further show that output-access protocol matters. A table-of-contents interface can reduce token cost while preserving or improving accuracy for stronger models, but it degrades performance for mid-tier models that cannot reliably navigate structured simulator outputs. PHREEQC-MCQ-200 therefore frames scientific tool use as an end-to-end diagnostic problem rather than a simple tool-calling capability. We argue that evaluations of scientific agents should report not only accuracy, but also item-level retention, output-access sensitivity, trajectory failures, and where the computation chain breaks.

cs.AI↗

arXiv2026

NewPINNs: Physics-Informing Neural Networks Using Conventional Solvers for Partial Differential Equations

We introduce NewPINNs, a physics-informing learning framework that couples neural networks with conventional numerical solvers for solving differential equations. Rather than enforcing governing equations and boundary conditions through residual-based loss terms, NewPINNs integrates the solver directly into the training loop and defines learning objectives through solver-consistency. The neural network produces candidate solution states that are advanced by the numerical solver, and training minimizes the discrepancy between the network prediction and the solver-evolved state. This pull-push interaction enables the network to learn physically admissible solutions through repeated exposure to the solver's action, without requiring problem-specific loss engineering or explicit evaluation of differential equation residuals. By delegating the enforcement of physics, boundary conditions, and numerical stability to established numerical solvers, NewPINNs mitigates several well-known failure modes of standard physics-informed neural networks, including optimization pathologies, sensitivity to loss weighting, and poor performance in stiff or nonlinear regimes. We demonstrate the effectiveness of the proposed approach across multiple forward and inverse problems involving finite volume, finite element, and spectral solvers.

cs.LG↗

arXiv2026

Understanding Tool-Augmented Agents for Lean Formalization: A Factorial Analysis

Automatic translation of natural language mathematics into faithful Lean 4 code is hindered by the fundamental dissonance between informal set-theoretic intuition and strict formal type theory. This gap often causes LLMs to hallucinate non-existent library definitions, resulting in code that fails to compile or lacks semantic fidelity. In this work, we investigate the effectiveness of tool-augmented agents for this task through a systematic factorial analysis of three distinct tool categories: Fine-tuned Model Querying (accessing expert drafts), Knowledge Search (retrieving symbol definitions), and Compiler Feedback (verifying code via a Lean REPL). We first benchmark the agent against one-shot baselines, demonstrating large gains in both compilation success and semantic equivalence. We then use the factorial decomposition to quantify the impact of each category, isolating the marginal contribution of each tool type to overall performance.

cs.SE↗

arXiv2026

Learning Parameterized Nonlinear Elasticity on Curved Surfaces

We learn parameterized nonlinear elasticity on curved surfaces using a physics-informed neural network that enforces governing equations and boundary conditions directly through the loss function, enabling a single trained model to represent a continuous family of elastic equilibria across geometric and material parameters. Nonlinear elasticity on curved manifolds underlies the mechanics of crystalline shells, elastic membranes, and viral capsids, where curvature and topological defects determine equilibrium structure and stability. Traditional exact and finite element solvers rely on symmetry reduction and must be reinitialized for each parameter choice, limiting scalability when symmetry is broken or parameters vary. We validate the proposed learning-based solver on a benchmark problem from curved elasticity, namely the one-dimensional single disclination on a spheroidal surface with known exact and numerical solutions. The network accurately reproduces these solutions, including parameter combinations excluded from training, demonstrating generalization across geometry and material regimes. This study establishes a scalable framework for learning nonlinear elastic systems on curved manifolds and lays the groundwork for extensions to fully two-dimensional and multi-defect configurations relevant to protein shells and other curved elastic networks.

physics.bio-ph↗

arXiv2026

PUNCH: Physics-informed Uncertainty-aware Network for Coronary Hemodynamics

More than 10 million coronary angiograms are performed globally each year, providing a gold standard for detecting obstructive coronary artery disease. Yet, no obstructive lesions are identified in 70% of patients evaluated for ischemic heart disease. Up to half of these patients have undiagnosed, life-limiting coronary microvascular dysfunction (CMD), which remains under-detected due to the limited availability of invasive tools required to measure coronary flow reserve (CFR). Here, we introduce PUNCH, a non-invasive, uncertainty-aware framework for estimating CFR directly from standard coronary angiography. PUNCH integrates physics-informed neural networks with variational inference to infer coronary blood flow from first-principles models of contrast transport, without requiring ground-truth flow measurements or population-level training. The pipeline runs in approximately three minutes per patient on a single GPU. Validated on synthetic angiograms with controlled noise and imaging artifacts, as well as on clinical bolus thermodilution data from 20 patients, PUNCH demonstrates accurate and uncertainty-calibrated CFR estimation. This approach establishes a new paradigm for CMD diagnosis and illustrates how physics-informed inference can substantially expand the diagnostic utility of available clinical imaging.

cs.LG↗

arXiv2026

Test-Driven Agentic Framework for Reliable Robot Controller

In this work, we present a test-driven, agentic framework for synthesizing a deployable low-level robot controller for navigation tasks. Given a 2D map with an image of an ultrasonic sensor-based robot, or a 3D robotic simulation environment, our framework iteratively refines the generated controller code using diagnostic feedback from structured test suites to achieve task success. We propose a dual-tier repair strategy to refine the generated code that alternates between prompt-level refinement and direct code editing. We evaluate the approach across 2D navigation tasks and 3D navigation in the Webots simulator. Experimental results show that test-driven synthesis substantially improves controller reliability and robustness over one-shot controller generation, especially when the initial prompt is underspecified. The source code and demonstration videos are available at: https://shivanshutripath.github.io/robotic_controller.github.io.

cs.RO↗

arXiv2026

Data-Efficient Physics-Informed Learning to Model Synchro-Waveform Dynamics of Grid-Integrated Inverter-Based Resources

Inverter-based resources (IBRs) exhibit fast transient dynamics during network disturbances, which often cannot be properly captured by phasor and SCADA measurements. This shortcoming has recently been addressed with the advent of waveform measurement units (WMUs), which provide high-resolution, time-synchronized raw voltage and current waveform samples from multiple locations in the power system. However, transient model learning based on synchro-waveform measurements remains constrained by the scarcity of network disturbances and the complexity of the underlying nonlinear dynamics of IBRs. We propose to address these problems by developing a data-efficient physics-informed machine learning (PIML) framework for synchro-waveform analytics that estimates the IBR terminal current response from only a few network disturbance signatures. Here, the physics of the electrical circuits are used to compensate for limited data availability by constraining the learning process through known circuit relationships. Two cases are considered, with known and unknown circuit parameters. In the latter case, the framework jointly learns the transient dynamics of the IBRs and the parameters of the electrical circuit. Case studies using WMU disturbance data across multiple sampling rates shows consistently lower current estimation error with substantially fewer training events than a purely data-driven baseline.

eess.SP↗

arXiv2025

ELPINN: Eulerian Lagrangian Physics-Informed Neural Network

Physics-Informed Neural Networks (PINNs) have gained widespread popularity for solving inverse and forward problems across a range of scientific and engineering domains. However, most existing PINN frameworks are limited to the Eulerian domain, where physical quantities are described at fixed spatial locations. In this work, we propose a novel PINN-based framework that couples Eulerian and Lagrangian perspectives by using particle trajectory data to reconstruct Eulerian velocity and pressure fields. We evaluate the performance of our method across three distinct fluid flow scenarios: two-dimensional external flow past a cylinder, two-dimensional internal flow in a confined geometry, and three-dimensional internal flow inside an airplane cabin. In all three cases, we successfully reconstruct the velocity field from Lagrangian particle data. Moreover, for the 2D external and internal flows, we recover the pressure field solely through the physics-informed learning process, without using any direct pressure measurements. We also conduct a sensitivity analysis to understand the effects of temporal resolution and particle count on the reconstruction accuracy. Our results show that smaller time-step sizes significantly improve the predictions, while the total number of particles has a comparatively smaller influence. These findings establish the potential of our coupled Eulerian-Lagrangian PINN framework as a powerful tool for enhancing experimental methods such as Particle Tracking Velocimetry (PTV). Looking ahead, this approach may be extended to infer hidden quantities such as pressure in three-dimensional flows or material properties like viscosity, opening new avenues for data-driven fluid dynamics in complex geometries.

physics.flu-dyn↗

arXiv2025

Aligning to What? Limits to RLHF Based Alignment

Reinforcement Learning from Human Feedback (RLHF) is increasingly used to align large language models (LLMs) with human preferences. However, the effectiveness of RLHF in addressing underlying biases remains unclear. This study investigates the relationship between RLHF and both covert and overt biases in LLMs, particularly focusing on biases against African Americans. We applied various RLHF techniques (DPO, ORPO, and RLOO) to Llama 3 8B and evaluated the covert and overt biases of the resulting models using matched-guise probing and explicit bias testing. We performed additional tests with DPO on different base models and datasets; among several implications, we found that SFT before RLHF calcifies model biases. Additionally, we extend the tools for measuring biases to multi-modal models. Through our experiments we collect evidence that indicates that current alignment techniques are inadequate for nebulous tasks such as mitigating covert biases, highlighting the need for capable datasets, data curating techniques, or alignment tools.

cs.CL↗

arXiv2025

Fine Tuning without Catastrophic Forgetting via Selective Low Rank Adaptation

Adapting deep learning models to new domains often requires computationally intensive retraining and risks catastrophic forgetting. While fine-tuning enables domain-specific adaptation, it can reduce robustness to distribution shifts, impacting out-of-distribution (OOD) performance. Pre-trained zero-shot models like CLIP offer strong generalization but may suffer degraded robustness after fine-tuning. Building on Task Adaptive Parameter Sharing (TAPS), we propose a simple yet effective extension as a parameter-efficient fine-tuning (PEFT) method, using an indicator function to selectively activate Low-Rank Adaptation (LoRA) blocks. Our approach minimizes knowledge loss, retains its generalization strengths under domain shifts, and significantly reduces computational costs compared to traditional fine-tuning. We demonstrate that effective fine-tuning can be achieved with as few as 5\% of active blocks, substantially improving efficiency. Evaluations on pre-trained models such as CLIP and DINO-ViT demonstrate our method's broad applicability and effectiveness in maintaining performance and knowledge retention.

cs.CV↗

arXiv2024

Where Did Your Model Learn That? Label-free Influence for Self-supervised Learning

Self-supervised learning (SSL) has revolutionized learning from large-scale unlabeled datasets, yet the intrinsic relationship between pretraining data and the learned representations remains poorly understood. Traditional supervised learning benefits from gradient-based data attribution tools like influence functions that measure the contribution of an individual data point to model predictions. However, existing definitions of influence rely on labels, making them unsuitable for SSL settings. We address this gap by introducing Influence-SSL, a novel and label-free approach for defining influence functions tailored to SSL. Our method harnesses the stability of learned representations against data augmentations to identify training examples that help explain model predictions. We provide both theoretical foundations and empirical evidence to show the utility of Influence-SSL in analyzing pre-trained SSL models. Our analysis reveals notable differences in how SSL models respond to influential data compared to supervised models. Finally, we validate the effectiveness of Influence-SSL through applications in duplicate detection, outlier identification and fairness analysis. Code is available at: \url{https://github.com/cryptonymous9/Influence-SSL}.

cs.LG↗