arXiv:2605.15775v1 Announce Type: new Abstract: Continual learning (CL) aims to train models sequentially over multiple domains without forgetting previously learned knowledge. However, existing CL methods optimize for in-domain performance and are therefore prone to learning spurious, domain-specific cues (``shortcut learning''), which limits generalization to unseen domains after deployment. In this paper, we address this limitation through continual learning of domain-invariant representation. We introduce a broad class of CL methods that sequentially learn representations capturing invariant structures across domains. Our methods are motivated by the observation that such invariant structures often preserve the underlying causal mechanisms, which can reduce the risk of overfitting to domain-specific cues and thus offer better out-of-domain generalization. Our proposed CL methods combine replay-based training with a tailored sequential invariance alignment to learn -- and preserve -- invariant structures over time. We evaluate our methods under a deployment-oriented protocol that measures performance on unseen target domains. Across six benchmark and real-world datasets spanning vision, medicine, manufacturing, and ecology, our methods consistently outperform existing CL baselines in terms of generalization to unseen target domains. As an ablation, we further show that na\"ive extensions of sequential training with existing domain-invariant representation learning (DIRL) methods provide only limited benefits. To the best of our knowledge, this is the first work to develop domain-invariant representation methods for CL.
Science Journals
arXiv:2505.21698v3 Announce Type: replace Abstract: Vision-language foundation models achieve promising performance in natural image classification, yet their direct application to medical imaging is limited by severe domain shifts, resolution mismatches, and the multi-label nature of clinical diagnosis. Training dedicated medical foundation models from scratch, however, is costly and data-intensive. Here, we propose MedBridge, a lightweight adaptation framework that opens a new direction in domain-gap mitigation by jointly combining domain alignment, resolution preservation, and multi-label reasoning via complementary VLM experts for medical image diagnosis. Specifically, MedBridge transforms pretrained VLMs into multi-view query encoders that inject a compact set of learnable query tokens into intermediate layers, enabling non-destructive domain alignment while preserving fine-grained pathological cues via multi-view high-resolution sampling. These query tokens further act as routing signals for a mixture-of-experts, dynamically integrating heterogeneous foundation models for multi-label reasoning without requiring a shared representation space. We evaluated MedBridge on five chest radiograph benchmarks in three key adaptation tasks. MedBridge demonstrates superior performance in both cross-domain generalization (out-of-distribution transfer) and in-domain specialization (same-distribution tuning) settings, yielding a significant 6-15% AUC improvement over state-of-the-art adaptation methods for multi-label thoracic disease diagnosis. Furthermore, MedBridge is model-agnostic and demonstrates broad extensibility across eight diverse VLMs (e.g., CLIP, LLaVA, Qwen-VL, MedGemma), highlighting its ability to flexibly adapt arbitrary foundation models into a powerful medical diagnostic tool. Our code will be released upon acceptance.
arXiv:2605.15700v1 Announce Type: new Abstract: Automated machine learning pipelines increasingly produce models whose predictions must be explained to end users, auditors, and downstream decision systems. The most widely used feature attribution methods (SHAP, Integrated Gradients, LIME) are typically chosen by convention rather than measured fidelity, because rigorous evaluation is impeded by the absence of ground-truth attribution on real data. We propose AGOP-IxG, a fast per-sample attribution method for tabular classifiers that pre-multiplies the per-sample gradient by a top-$K$ rank-truncated Average Gradient Outer Product matrix, and evaluate it against four widely-used baselines on a controlled tabular benchmark designed for AutoML practitioners. In Part 1, we construct three synthetic multi-class tabular tasks (linear, sparse nonlinear, interaction-based) where ground-truth attribution per sample is analytically or numerically derivable, and compare five methods: AGOP-IxG, SHAP (DeepExplainer), Integrated Gradients, InputXGradient, and LIME. AGOP-IxG leads on Spearman rank correlation and noise feature mass on all three synthetic datasets, and on top-$k$ precision on the interaction dataset. Across all settings, AGOP-IxG is approximately $350\times$ to $1{,}650\times$ faster than SHAP. In Part 2, we evaluate global faithfulness on Adult Income and Credit Card Default using the ROAR protocol; the methods cluster within $\sim 1.7\%$ relative AUC, consistent with AGOP-IxG being optimized for per-sample local attribution rather than global feature ranking.
arXiv:2605.15664v1 Announce Type: new Abstract: Non-wellfounded proof systems impose a global condition called the global trace condition (GTC) on a derivation tree to ensure soundness. Providing a categorical characterisation of the GTC that guarantees soundness remains challenging due to the global, non-compositional nature of these conditions and the infinitary structure of non-wellfounded proofs. We develop a coalgebraic framework for non-wellfounded proof systems where derivation trees are modelled as coalgebras of generalised polynomial functors on presheaves. Since the GTC is a constraint on infinite paths in derivation graphs, we employ graphs of coalgebras and formulate the GTC coalgebraically as a condition on these graphs. Soundness is then formulated as the existence of a unique coalgebra-to-algebra morphism from a coalgebra representing a derivation graph to an algebra specifying semantics. Within this framework, we characterise the GTC via recursive coalgebras: a coalgebra satisfies the GTC if and only if its image under a suitable adjoint is recursive. Under an appropriate assumption on the given semantic algebra, this yields soundness, that is, every proof admits a unique coalgebra-to-algebra morphism. We demonstrate our framework through a non-wellfounded proof system for the modal mu-calculus, one for higher-order fixed-point logics, and a non-wellfounded variant of Santocanale's circular proof system in mu-bicomplete categories.
arXiv:2409.03897v3 Announce Type: replace Abstract: Large-scale multi-agent systems are often deployed across wide geographic areas, where agents interact with heterogeneous environments. There is an emerging interest in understanding the role of heterogeneity in the performance of the federated versions of classic reinforcement learning algorithms. In this paper, we study synchronous federated Q-learning, which aims to learn an optimal Q-function by having $K$ agents average their local Q-estimates per $E$ iterations. We observe an interesting phenomenon on the convergence speeds in terms of $K$ and $E$. Similar to the homogeneous environment settings, there is a linear speed-up concerning $K$ in reducing the errors that arise from sampling randomness. Yet, in sharp contrast to the homogeneous settings, $E>1$ leads to significant performance degradation. Specifically, we provide a fine-grained characterization of the error evolution in the presence of environmental heterogeneity, which decay to zero as the number of iterations $T$ increases. The slow convergence of having $E>1$ turns out to be fundamental rather than an artifact of our analysis. We prove that, for a wide range of stepsizes, the $\ell_{\infty}$ norm of the error cannot decay faster than $\Theta (E/T)$. In addition, our experiments demonstrate that the convergence exhibits an interesting two-phase phenomenon. For any given stepsize, there is a sharp phase-transition of the convergence: the error decays rapidly in the beginning yet later bounces up and stabilizes. Provided that the phase-transition time can be estimated, choosing different stepsizes for the two phases leads to faster overall convergence.
arXiv:2601.00678v3 Announce Type: replace Abstract: Humans excel at forecasting the future dynamics of a scene given just a single image. Video generation models that can mimic this ability are an essential component for intelligent systems. Recent approaches have improved temporal coherence and 3D consistency in single-image-conditioned video generation. However, these methods often lack robust user controllability, such as modifying the camera path, limiting their applicability in real-world applications. Most existing camera-controlled image-to-video models struggle with accurately modeling camera motion, maintaining temporal consistency, and preserving geometric integrity. Leveraging explicit intermediate 3D representations offers a promising solution by enabling coherent video generation aligned with a given camera trajectory. Although these methods often use 3D point clouds to render scenes and introduce object motion in a later stage, this two-step process still falls short in achieving full temporal consistency, despite allowing precise control over camera movement. We propose a novel framework that constructs a 3D Gaussian scene representation and samples plausible object motion, given a single image in a single forward pass. This enables fast, camera-guided video generation without the need for iterative denoising to inject object motion into render frames. Extensive experiments on the KITTI, Waymo, RealEstate10K and DL3DV-10K datasets demonstrate that our method achieves state-of-the-art video quality and inference efficiency. The project page is available at https://melonienimasha.github.io/Pixel-to-4D-Website.
arXiv:2605.15467v1 Announce Type: new Abstract: Conversational nurse-patient transcripts contain actionable observations, but converting these transcripts into structured representations at scale remains challenging. Documentation burden is substantial, with prior studies showing clinicians spend large portions of their workday on documentation and related desk work rather than direct patient care. MEDIQA-SYNUR focuses on observation extraction from conversational nurse-patient transcripts, requiring systems to normalize these narratives into a predefined schema with value-type constraints. We propose a modular retrieval-augmented generation (RAG) pipeline that uses the training set as an exemplar corpus, combines schema-constrained prompting (full schema vs. pruned candidate schema), deterministic schema-based postprocessing, and a second-pass audit, with two LLM backbones: Llama-4-Scout-17B-16E-Instruct and GPT-5.2 with corresponding embedding models for RAG. Our best configuration uses GPT-5.2 with full schema, RAG, and a second-pass auditing, achieving 80.36% F1 score. Overall, our results show that RAG consistently improves performance, while the optimal degree of schema constraint depends on the model, and second-pass auditing yields modest additional gains by correcting residual schema-adherence errors.
arXiv:2605.15698v1 Announce Type: new Abstract: Recent attempts at creating Foundation Models (FMs) for Electroencephalography (EEG) have achieved state-of-the-art performance on multiple tasks including Motor Imagery (MI). These MI tasks have typically involved coarse classification between imagined limb movements. However, the development of foundation models necessitates diverse datasets, both for pretraining and evaluating the progress of these models. In this work, we propose handwriting decoding as a challenging motor task for FMs. We show that several existing datasets are potentially confounded, and introduce a dataset that more rigorously evaluates models. On this dataset, we find that current FMs, despite showing SOTA performance in multiple MI datasets are outperformed by smaller task-specific models. We also highlight challenges specific to EEG-based handwriting decoding to inform future work. In our 4-letter classification task, we show that (a) Knowledge of movement-onset is crucial to reported decoding performance in prior works, with average performance across subjects dropping from $41.3\%$ to $32.4\%$. (b) Increasing test-time signal quality provides significant performance improvements ($45\%$ to $78\%$ in our best subject) compared to scaling training data with single-trial EEG. (c) While scaling training data steadily improves decoding performance, existing FMs do not outperform specialist models in handwriting decoding. We make our code available at https://anonymous.4open.science/r/EEG-Handwriting-BCI-DFCD/
arXiv:2605.15790v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) improves reliability of large language models by incorporating external knowledge, but the retrieval process can introduce bias that propagates to generated outputs. This issue is particularly challenging in top-k settings, where multiple documents jointly influence generation. We propose a fairness-aware retrieval framework that models and controls this bias. Our approach combines controlled bias injection via reranking, a position-aware model of bias propagation, and an optimization formulation that balances relevance and fairness. We further introduce a scalable solution based on Quadratic Fairness via Dual Hyperplane Approximation (FARO), which enables efficient optimization through problem decomposition. Experimental results show that our method effectively mitigates generation bias while preserving relevance. This work provides a principled approach for fairness-aware retrieval in RAG systems.
arXiv:2605.15851v1 Announce Type: new Abstract: This paper presents a data-driven framework for uncertainty propagation under unmeasured or statistically unmodeled (unstructured) disturbances. We consider residual disturbances, which consolidate all unstructured disturbances into a single quantity that can be estimated from data. Under mild assumptions, the resulting stochastic predictor is causal and distributionally consistent, enabling efficient uncertainty quantification through polynomial chaos expansions and higher-order Chebyshev inequalities. The proposed method is validated using experimental data from a smart home in Norway.
arXiv:2605.15789v1 Announce Type: new Abstract: Uncertainty quantification is essential in safety-critical settings--from autonomous driving to aviation, finance, and health--where decisions must rely on conservative bounds rather than point estimates. Predictor-level intervals (e.g., from quantile regression, conformal prediction, variance networks, or Bayesian models) generally do not compose: adding two per-variable intervals need not yield a valid interval for their sum or preserve coverage. In aviation, Gaussian overbounding replaces complex error distributions with a conservative Gaussian whose tails dominate the truth, so conservatism propagates through linear operations. Yet classical overbounds are global, often overly conservative, and hard to adapt to feature-conditioned errors. We propose a unified learning framework that trains neural networks to produce context-aware Gaussian overbounds--mean and scale--with provable conservatism on a finite quantile grid and, under three explicit regularity assumptions, continuous-tail conservatism on a certified interval. Our overbounding loss enforces conservativeness at selected quantiles while penalizing distributional distance with a Wasserstein-style term. The learned bounds support conservative linear-combination and convolution analysis on the enforced grid, and on the certified interval when assumptions hold, while being less redundant than traditional methods. We provide a scoped analysis of discrete-to-continuous conservatism and compact-domain objective regularity, and validate on synthetic data and real-world datasets, including multipath, ionospheric, and tropospheric residual errors. Across these settings, the method yields tighter bounds while maintaining conservatism on the enforced grid and in experiments. The framework is modality-agnostic and applicable to learning systems that require conservative, feature-conditioned uncertainty estimates in dynamic environments.
arXiv:2605.15782v1 Announce Type: new Abstract: In this work, we address the problem of ensuring real-time safety in autonomous robot navigation, in spatially constrained dynamic environments, by utilizing only onboard sensors. We present a real-time control architecture that integrates a 3D LIDAR perception-based composite control barrier function(CBF)-based safety filter directly into the autonomy pipeline. The proposed perception-driven framework enforces collision avoidance constraints dynamically from onboard point cloud data, thus allowing a large number of constraints to be handled at the control frequency, while remaining minimally invasive to nominal task execution. The safety region is defined as an ellipsoid in the body-frame, consistent with the geometry of the platform, which induces time-varying constraints in the world frame as the robot rotates; this effect is handled through a dedicated formulation of time-varying (CBF) for each LIDAR point. We validate the system through multiple field experiments in underground environments by utilizing a quadruped platform performing a visual inspection task, demonstrating reliable operation in the presence of dynamic obstacles, unsafe high-level references, abrupt localization anomalies, and while traversing through narrow corridors.
arXiv:2509.01685v2 Announce Type: replace-cross Abstract: We consider sampling from a Gibbs distribution by evolving finitely many particles. We propose a preconditioned version of a recently proposed noise-free sampling method, governed by approximating the score function with the numerically tractable score of a regularized Wasserstein proximal operator. This is derived by a Cole--Hopf transformation on coupled anisotropic heat equations, yielding a kernel formulation for the preconditioned regularized Wasserstein proximal. The diffusion component of the proposed method is also interpreted as a modified self-attention block, as in transformer architectures. For quadratic potentials, we provide a discrete-time non-asymptotic convergence analysis and explicitly characterize the bias, which is dependent on regularization and independent of step-size. Experiments demonstrate acceleration and particle-level stability on various log-concave and non-log-concave toy examples to Bayesian total-variation regularized image deconvolution, and competitive/better performance on non-convex Bayesian neural network training when utilizing variable preconditioning matrices.
arXiv:2605.15848v1 Announce Type: new Abstract: Conversational interfaces powered by large language models (LLMs) are widely used for ideation and analysis, yet their linear structure limits exploration of alternatives and management of long-running interactions. We present CanvasConvo, a conversational interface concept that transforms linear chat into a branching conversation tree embedded in a spatial canvas. CanvasConvo enables users to explore what-if scenarios by branching directly from conversational content, supporting parallel development of alternative directions. These branches are visualized on a canvas while remaining integrated with a familiar chat interface, allowing users to switch between linear and non-linear interaction. Features such as timeline-based navigation, automatic tagging and summarization, and context-aware controls (e.g., goals, reusable prompts) support structured interaction and continuity. We evaluated CanvasConvo in a 5-7 day field study with 24 participants. Our findings highlight how non-linear conversational structures support exploratory workflows and different interactions in LLM-based work.
arXiv:2511.13108v4 Announce Type: replace Abstract: The rapid progress of generative models such as GANs and diffusion models has led to the widespread proliferation of AI-generated images, raising concerns about misinformation, privacy violations, and trust erosion in digital media. Although large-scale multimodal models like CLIP offer strong transferable representations for detecting synthetic content, fine-tuning them often induces catastrophic forgetting, which degrades pre-trained priors and limits cross-domain generalization. To address this issue, we propose the Distillation-guided Gradient Surgery Network (DGS-Net), a novel framework that preserves transferable pre-trained priors while suppressing task-irrelevant components. Specifically, we introduce a gradient-space decomposition that separates harmful and beneficial descent directions during optimization. By projecting task gradients onto the orthogonal complement of harmful directions and aligning with beneficial ones distilled from a frozen CLIP encoder, DGS-Net achieves unified optimization of prior preservation and irrelevant suppression. Extensive experiments on 50 generative models demonstrate that our method outperforms state-of-the-art approaches by an average margin of 6.6, achieving superior detection performance and generalization across diverse generation techniques.
arXiv:2605.15714v1 Announce Type: new Abstract: This position paper argues that the machine learning community should prioritize early-stage quality assurance in annotation pipelines over the prevailing practice of late-stage validation. Data quality bottlenecks increasingly limit foundation model improvement, yet quality assurance research focuses almost exclusively on validation methods rather than validation timing. When validation occurs, not merely what methods are employed, fundamentally determines both error rates and annotation costs. This temporal neglect is puzzling given the well-established "shift-left" principle from software engineering, where empirical studies demonstrate 4--100x cost multipliers for defects detected in later stages (Boehm, 1981; Shull et al., 2002). Annotation pipelines exhibit analogous dynamics: errors caught before annotation begins cost a fraction of those discovered after review cycles complete. We propose a taxonomy of three QA trigger points, namely pre-annotation (T0), post-annotation (T1), and post-review (T2), that decompose annotation workflows into discrete validation opportunities. A parametric error-propagation model formalizes when timing affects final error rates versus only economics, making timing a measurable design variable rather than a configuration afterthought. A survey of 47 recent papers reveals that only 4% report when validation occurs, a striking gap given timing's demonstrated impact in adjacent fields. Without explicit attention to QA timing, the community risks optimizing validation methods while ignoring the structural variable that may matter most. Acting on this position requires three steps: researchers should report QA timing configurations alongside validation methods; annotation platforms should expose timing as a first-class parameter; and the community should run controlled experiments that measure stage-specific detection rates directly.
arXiv:2605.00934v2 Announce Type: replace Abstract: Coherent Point Drift (CPD) is a representative probabilistic framework for unsupervised non-rigid point set registration. Its standard non-rigid M-step, however, relies on a point-indexed Gaussian-kernel system whose size grows with the number of moving points, making deformation estimation computationally heavy for large point sets and difficult to control in complexity during registration. To address these limitations, we propose Analytic-CPD, a new unsupervised non-rigid registration framework that gives CPD a structured analytic reformulation. Analytic-CPD preserves the CPD posterior correspondence layer, but lifts the M-step from point-indexed kernel displacement estimation to structured analytic mapping estimation. By coupling the Gaussian-mixture posterior mechanism of CPD with Structured Analytic Mappings (SAM), the method obtains a deformation model whose coefficient dimension is governed by the ambient dimension and analytic order rather than by the number of moving points. More importantly, deformation estimation is organized over an interpretable hierarchy of analytic function spaces, so the analytic order can be increased progressively as posterior correspondences become more reliable. We implement this idea through an increasing-degree continuation strategy with decreasing stage lengths: low-order analytic maps first stabilize the posterior correspondence structure, while higher-order modes later refine nonlinear residual deformation. Experiments on controlled model-matched, smooth model-mismatch, and registered human-shape data demonstrate the effectiveness and favorable accuracy--efficiency performance of Analytic-CPD.
arXiv:2605.16008v1 Announce Type: new Abstract: Plaque assays remain the gold standard readout of virus infectivity; however, plaque counting from plate images is labor-intensive and prone to inter-operator variability. We present an end-to-end, computer-aided workflow for cytopathic effect-based virus titration directly from laboratory plaque assay images. The proposed approach combines two models derived from the Segment Anything Model (SAM): a SAM2-based well-segmentation module that localizes assay wells across heterogeneous imaging conditions, and a SAM-based plaque-segmentation model that detects and enumerates plaques within each well. The method was evaluated on a mixed dataset comprising private plaque assay images of Mayaro virus and Coxsackievirus B3, together with public Vaccinia virus images from the VACVPlaque dataset. The pipeline outputs per-well plaque counts, automatically computes plaque-forming units per milliliter (PFU/mL), and is integrated into a web-based platform that allows users to review results and organize experiments. On held-out plates (17 from MAYV/CVB3 and 22 from VACV), the workflow generalized across two plate formats (6-well and 12-well) and showed strong agreement with manual annotations (Pearson correlation coefficients of 0.92 for MAYV/CVB3 and 0.88 for VACV). Automated plaque counts were further compared with annotations from four independent experts, demonstrating high concordance. The proposed system will be open sourced and publicly released upon acceptance of this manuscript to enable reproducible, scalable, and audit-ready plaque assay analysis while substantially reducing manual annotation effort.
arXiv:2505.18511v3 Announce Type: replace Abstract: Stochastic Partial Differential Equations (SPDEs) driven by random noise play a central role in modeling physical processes with rough spatio-temporal dynamics, such as turbulence flows, superconductors, and quantum dynamics. Although machine learning (ML)-based surrogate models have shown promise for efficiently approximating such dynamics, progress remains limited by the lack of a unified benchmark with controlled data generation and comprehensive evaluation. This gap is particularly significant for singular SPDEs, for which benchmark datasets are largely unavailable and reliable simulation requires numerically delicate schemes based on renormalization. Moreover, subtle differences in data-generation procedures, such as noise approximation, basis choice, and the inclusion of renormalization, can significantly affect the resulting datasets and, consequently, model evaluation. We introduce SPDEBench, the first unified benchmark for ML-based SPDE learning. SPDEBench provides ready-to-use datasets for physically and mathematically significant SPDEs on 1-3D domains with periodic or Dirichlet boundary condition. Both regular and singular SPDEs are taken into consideration. SPDEBench also incorporates representative ML baselines in operator learning, together with 7 evaluation metrics, including Sobolev and distributional metrics beyond the standard $L^2$-error. Supported by SPDEBench, we conduct systematic evaluations of model accuracy, robustness, and out-of-distribution generalization under controlled data variations. Our numerical results show that SPDE-aware architectures generally achieve stronger performance than generic operator-learning baselines. These findings establish SPDEBench as a reproducible and extensible resource, paving pathway for principled benchmarking and architecture design for stochastic spatio-temporal dynamics.
arXiv:2603.25099v2 Announce Type: replace Abstract: We present a framework in which a large language model (LLM) acts as an online adaptive controller for SIMP topology optimization, replacing conventional fixed-schedule continuation with real-time, state-conditioned parameter decisions. At every $k$-th iteration, the LLM receives a structured observation$-$current compliance, grayness index, stagnation counter, checkerboard measure, volume fraction, and budget consumption$-$and outputs numerical values for the penalization exponent $p$, projection sharpness $\beta$, filter radius $r_{\min}$, and move limit $\delta$ via a Direct Numeric Control interface. A hard grayness gate prevents premature binarization, and a meta-optimization loop uses a second LLM pass to tune the agent's call frequency and gate threshold across runs. We benchmark the agent against four baselines$-$fixed (no-continuation), standard three-field continuation, an expert heuristic, and a schedule-only ablation$-$on three 2-D problems (cantilever, MBB beam, L-bracket) at $120\!\times\!60$ resolution and two 3-D problems (cantilever, MBB beam) at $40\!\times\!20\!\times\!10$ resolution, all run for 300 iterations. A standardized 40-iteration sharpening tail is applied from the best valid snapshot so that compliance differences reflect only the exploration phase. The LLM agent achieves the lowest final compliance on every benchmark: $-5.7\%$ to $-18.1\%$ relative to the fixed baseline, with all solutions fully binary. The schedule-only ablation underperforms the fixed baseline on two of three problems, confirming that the LLM's real-time intervention$-$not the schedule geometry$-$drives the gain. Code and reproduction scripts will be released upon publication.
arXiv:2605.16009v1 Announce Type: new Abstract: Local navigation is one of the fundamental problems in robot navigation, and numerous approaches have been proposed over the years, including methods such as the Dynamic Window Approach, Model Predictive Control, and more recently, Control Barrier Functions and machine learning based techniques. While these methods perform well in simple environments, many of them rely on optimization or learning based procedures that can struggle in more complex scenarios. In contrast, this article proposes a more geometric algorithmic approach that enables a local navigation method with faster computation times and longer planning horizons. The proposed method is based on the computation of a sequence of circular regions from a local LiDAR scan that expand in the direction of the goal and capture free local navigable space. The proposed method was implemented in the ROS2 framework and evaluated in a simulated environment.
arXiv:2605.15464v1 Announce Type: new Abstract: Post-training has become a crucial step for unlocking the capabilities of large language models, with reinforcement learning (RL) emerging as a critical paradigm. Recent RL-based post-training has increasingly split into two paradigms: reinforcement learning from human feedback (RLHF), which optimizes models using human preference signals in target domains, and reinforcement learning from verifiable rewards (RLVR), which operates in verifier-backed environments. The latter has dominated recent reasoning-oriented post-training because it delivers stronger gains and higher efficiency on domain-specific tasks (e.g., reasoning). However, although in-domain RL training achieves promising performance, it still requires a substantial amount of GPU compute, which remains a major barrier to broad adoption. In this work, we study the generalization ability of RLHF learned from scratch from a small set of interactions in open-ended environments, and investigate whether the conversational abilities it explicitly acquires can implicitly transfer to downstream tasks such as mathematical reasoning and code generation, namely GRLO. Specifically, on Qwen3-4B-Base backbone, GRLO improves the average performance across all domains from 24.1 to 63.1 with only 5K prompts and 22.7 GPU hours, requiring about $46\times$ less data and $68\times$ less compute than a strong in-domain RLVR baseline. The resulting model is even competitive with Qwen's released post-trained models which required a much larger training cost. Notably, a subsequent in-domain RLVR stage brings only selective gains, mainly on harder competition-math benchmarks. We hope GRLO offers a simple and efficient recipe for building broadly capable post-trained models. Our code and data will be available at: \href{https://github.com/SJY8460/GRLO}{https://github.com/SJY8460/GRLO}.
arXiv:2605.16011v1 Announce Type: new Abstract: Adaptive learning refers to educational technologies that track learners' learning progress and adapt the instructional process based on individual learners' learning performance. It is increasingly recognized as critical for developing an effective learning support tool. Vision language models (VLMs) have seen adoption in mathematics education, and students have been using them as learning aids for personalized instruction. However, it is unknown whether VLMs have the ability to adapt to different learner profiles when providing mathematical instructions. Current VLMs lack a systematic evaluation framework for this adaptivity to different learner profiles in mathematics tutoring tasks. To address this gap, we draw on the learner model from the adaptive learning framework (Shute and Towle, 2018) and propose a learner model-based rubric. Our rubric formalizes adaptivity assessment into three aspects: cognitive aspects, motivational aspects, and complexity. We also evaluate two additional dimensions of VLM responses: correctness (of answers and solutions) and quality (of the response itself). Our experimental results show measurable differences in adaptivity across models and also reveal that current VLMs struggle to consistently produce learner model-based instructional responses, especially when receiving limited learner information.
arXiv:2605.15617v1 Announce Type: new Abstract: Large language model (LLM) training today runs on clusters spanning thousands of GPUs. While this scale enables rapid model advances, developing, debugging, and performance-tuning the training framework inevitably becomes complex and costly. This is because engineers often need to reproduce production behaviors to diagnose failures or evaluate optimizations, thereby demanding frequent and even exclusive access to production-scale clusters -- which becomes increasingly hard given that the majority of GPUs are already committed to production workloads. Simulation relies on complex performance models that are difficult to maintain, and downscaled experiments often fail to capture scale-dependent behaviors. We present PrismLLM to decouple large-scale execution from the need to access large clusters, enabling engineers to run and observe ranks of interest under faithful large-scale behavior using only a few GPUs. PrismLLM constructs a high-fidelity execution graph via a slicing-based approach that captures computation, communication, and dependencies of the target scale. Then, PrismLLM performs hybrid emulation where selected ranks execute the original program while the remaining ranks are replayed as virtual participants. Experiments on large-scale LLM training workloads show that PrismLLM accurately reproduces performance and memory behavior, achieving only 0.58\% average error in iteration time and less than 0.01\% error in peak GPU memory usage. PrismLLM can emulate clusters of up to 8192 GPUs using fewer than 1\% of the physical GPUs required by the original deployment.
arXiv:2605.15463v1 Announce Type: new Abstract: As machine learning models grow in complexity, they increasingly struggle with three conflicting demands: the need for high accuracy, the requirement for hardware efficiency, and the necessity of functional stability. Traditional architectures often achieve performance at the expense of spiky or unpredictable behavior, where small changes in input lead to massive swings in output -- a critical flaw for real-world deployment in sensitive environments. This paper introduces ChainzRule (CR), a novel neural architecture designed to harmonize these competing goals. ChainzRule replaces standard piecewise-linear activations with a Polynomial Engine governed by Differential Regularization (DREG). Unlike traditional methods that impose global, coarse-grained constraints on a model's Lipschitz constant, DREG acts as a targeted regularization on intermediate derivatives. This approach suppresses extreme sensitivity without attenuating the representational power inherent in the Polynomial Engine. In head-to-head "Fair Fight" benchmarks, ChainzRule outperformed standard models while using 15.5x fewer parameters. On the MNIST dataset, it reduced peak gradient volatility by an average of 23.1%, ensuring a smoother and more predictable manifold. On Yelp Full ordinal regression under explicit DREG regularization, ChainzRule achieves 70.17% accuracy, validating that derivative-aware regularization is compatible with competitive performance on realistic tasks. By embedding gradient awareness into the architecture via DREG, ChainzRule demonstrates that stability and accuracy need not be competing objectives.