# Research Papers — 2026-03 · bundle page 76 of 243 > Full-text markdown of papers #3751–3800 of 12123 published in 2026-03, newest first. AI-readable mirrors of the full 12123-paper 2026-03 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-03/page/76 [← Previous bundle](https://aipod.dev/papers/archive/2026-03/page/75.md) [Next bundle →](https://aipod.dev/papers/archive/2026-03/page/77.md) ### Alignment Whack-a-Mole : Finetuning Activates Verbatim Recall of Copyrighted Books in Large Language Models Authors: Xinyue Liu, Niloofar Mireshghallah, Jane C. Ginsburg, Tuhin Chakrabarty Published: 2026-03-21 Categories: cs.CL, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2603.20957v5 Frontier LLM companies have repeatedly assured courts and regulators that their models do not store copies of training data. They further rely on safety alignment strategies via RLHF, system prompts, and output filters to block verbatim regurgitation of copyrighted works, and have cited the efficacy of these measures in their legal defenses against copyright infringement claims. We show that finetuning bypasses these protections: by training models to expand plot summaries into full text, a task naturally suited for commercial writing assistants, we cause GPT-4o, Gemini-2.5-Pro, and DeepSeek-V3.1 to reproduce up to 85-90% of held-out copyrighted books, with single verbatim spans exceeding 460 words, using only semantic descriptions as prompts and no actual book text. This extraction generalizes across authors: finetuning exclusively on Haruki Murakami's novels unlocks verbatim recall of copyrighted books from over 30 unrelated authors. The effect is not specific to any training author or corpus: random author pairs and public-domain finetuning data produce comparable extraction, while finetuning on synthetic text yields near-zero extraction, indicating that finetuning on individual authors' works reactivates latent memorization from pretraining. Three models from different providers memorize the same books in the same regions ($r \ge 0.90$), pointing to an industry-wide vulnerability. Our findings offer compelling evidence that model weights store copies of copyrighted works and that the security failures that manifest after finetuning on individual authors' works undermine a key premise of recent fair use rulings, where courts have conditioned favorable outcomes on the adequacy of measures preventing reproduction of protected expression. --- ### Alignment Whack-a-Mole : Finetuning Activates Verbatim Recall of Copyrighted Books in Large Language Models Authors: Xinyue Liu, Niloofar Mireshghallah, Jane C. Ginsburg, Tuhin Chakrabarty Published: 2026-03-21 Categories: cs.CL, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2603.20957v3 Frontier LLM companies have repeatedly assured courts and regulators that their models do not store copies of training data. They further rely on safety alignment strategies via RLHF, system prompts, and output filters to block verbatim regurgitation of copyrighted works, and have cited the efficacy of these measures in their legal defenses against copyright infringement claims. We show that finetuning bypasses these protections: by training models to expand plot summaries into full text, a task naturally suited for commercial writing assistants, we cause GPT-4o, Gemini-2.5-Pro, and DeepSeek-V3.1 to reproduce up to 85-90% of held-out copyrighted books, with single verbatim spans exceeding 460 words, using only semantic descriptions as prompts and no actual book text. This extraction generalizes across authors: finetuning exclusively on Haruki Murakami's novels unlocks verbatim recall of copyrighted books from over 30 unrelated authors. The effect is not specific to any training author or corpus: random author pairs and public-domain finetuning data produce comparable extraction, while finetuning on synthetic text yields near-zero extraction, indicating that finetuning on individual authors' works reactivates latent memorization from pretraining. Three models from different providers memorize the same books in the same regions ($r \ge 0.90$), pointing to an industry-wide vulnerability. Our findings offer compelling evidence that model weights store copies of copyrighted works and that the security failures that manifest after finetuning on individual authors' works undermine a key premise of recent fair use rulings, where courts have conditioned favorable outcomes on the adequacy of measures preventing reproduction of protected expression. --- ### Beyond Expression Similarity: Contrastive Learning Recovers Functional Gene Associations from Protein Interaction Structure Authors: Jason Dury Published: 2026-03-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.20955v1 The Predictive Associative Memory (PAM) framework posits that useful relationships often connect items that co-occur in shared contexts rather than items that appear similar in embedding space. A contrastive MLP trained on co-occurrence annotations--Contrastive Association Learning (CAL)--has improved multi-hop passage retrieval and discovered narrative function at corpus scale in text. We test whether this principle transfers to molecular biology, where protein-protein interactions provide functional associations distinct from gene expression similarity. Four experiments across two biological domains map the operating envelope. On gene perturbation data (Replogle K562 CRISPRi, 2,285 genes), CAL trained on STRING protein interactions achieves cross-boundary AUC of 0.908 where expression similarity scores 0.518. A second gene dataset (DepMap, 17,725 genes) confirms the result after negative sampling correction, reaching cross-boundary AUC of 0.947. Two drug sensitivity experiments produce informative negatives that sharpen boundary conditions. Three cross-domain findings emerge: (1) inductive transfer succeeds in biology--a node-disjoint split with unseen genes yields AUC 0.826 (Delta +0.127)--where it fails in text (+/-0.10), suggesting physically grounded associations are more transferable than contingent co-occurrences; (2) CAL scores anti-correlate with interaction degree (Spearman r = -0.590), with gains concentrating on understudied genes with focused interaction profiles; (3) tighter association quality outperforms larger but noisier training sets, reversing the text pattern. Results are stable across training seeds (SD < 0.001) and cross-boundary threshold choices. --- ### Before the Tool Call: Deterministic Pre-Action Authorization for Autonomous AI Agents Authors: Uchi Uchibeke Published: 2026-03-21 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2603.20953v1 AI agents today have passwords but no permission slips. They execute tool calls (fund transfers, database queries, shell commands, sub-agent delegation) with no standard mechanism to enforce authorization before the action executes. Current safety architectures rely on model alignment (probabilistic, training-time) and post-hoc evaluation (retrospective, batch). Neither provides deterministic, policy-based enforcement at the individual tool call level. We characterize this gap as the pre-action authorization problem and present the Open Agent Passport (OAP), an open specification and reference implementation that intercepts tool calls synchronously before execution, evaluates them against a declarative policy, and produces a cryptographically signed audit record. OAP enforces authorization decisions in a measured median of 53 ms (N=1,000). In a live adversarial testbed (4,437 authorization decisions across 1,151 sessions, $5,000 bounty), social engineering succeeded against the model 74.6% of the time under a permissive policy; under a restrictive OAP policy, a comparable population of attackers achieved a 0% success rate across 879 attempts. We distinguish pre-action authorization from sandboxed execution (contains blast radius but does not prevent unauthorized actions) and model-based screening (probabilistic), and show they are complementary. The same infrastructure that enforces security constraints (spending limits, capability scoping) also enforces quality gates, operational contracts, and compliance controls. The specification is released under Apache 2.0 (DOI: 10.5281/zenodo.18901596). --- ### gUFO: A Gentle Foundational Ontology for Semantic Web Knowledge Graphs Authors: João Paulo A. Almeida, Giancarlo Guizzardi, Tiago Prince Sales, Claudenir M. Fonseca Published: 2026-03-21 Categories: cs.AI, cs.DB arXiv: https://arxiv.org/abs/2603.20948v1 gUFO is a lightweight implementation of the Unified Foundational Ontology (UFO) suitable for Semantic Web OWL 2 DL applications. UFO is a mature foundational ontology with a rich axiomatization and that has been employed in a significant number of projects in research and industry. Moreover, it is currently in the process of standardization by the International Organization for Standardization as the ISO/IEC CD 21838-5. gUFO stands out from other foundational ontology implementations (such as those provided for BFO and DOLCE) given its unique support for a typology of types (operationalizing OntoClean guidelines), its reification patterns for intrinsic and relational aspects, and its support for situations and high-order types. gUFO provides well-founded patterns to address recurrent problems in Semantic Web knowledge graphs. In this paper, we present gUFO with its constituting categories, relations and constraints, discuss how it differs from the original UFO reference ontology, elaborate on its community adoption, and systematically position it in relation to existing OWL-based implementations of popular alternative foundational ontologies. --- ### User Preference Modeling for Conversational LLM Agents: Weak Rewards from Retrieval-Augmented Interaction Authors: Yuren Hao, Shuhaib Mehri, ChengXiang Zhai, Dilek Hakkani-Tür Published: 2026-03-21 Categories: cs.CL, cs.AI, cs.HC, cs.IR, stat.ML arXiv: https://arxiv.org/abs/2603.20939v1 Large language models are increasingly used as personal assistants, yet most lack a persistent user model, forcing users to repeatedly restate preferences across sessions. We propose Vector-Adapted Retrieval Scoring (VARS), a pipeline-agnostic, frozen-backbone framework that represents each user with long-term and short-term vectors in a shared preference space and uses these vectors to bias retrieval scoring over structured preference memory. The vectors are updated online from weak scalar rewards from users' feedback, enabling personalization without per-user fine-tuning. We evaluate on \textsc{MultiSessionCollab}, an online multi-session collaboration benchmark with rich user preference profiles, across math and code tasks. Under frozen backbones, the main benefit of user-aware retrieval is improved interaction efficiency rather than large gains in raw task accuracy: our full VARS agent achieves the strongest overall performance, matches a strong Reflection baseline in task success, and reduces timeout rate and user effort. The learned long-term vectors also align with cross-user preference overlap, while short-term vectors capture session-specific adaptation, supporting the interpretability of the dual-vector design. Code, model, and data are available at https://github.com/YurenHao0426/VARS. --- ### Two Approaches to Direct Estimation of Riesz Representers Authors: David Bruns-Smith Published: 2026-03-21 Categories: econ.EM, stat.ML arXiv: https://arxiv.org/abs/2603.20936v1 The Riesz representer is a central object in semiparametric statistics and debiased/doubly-robust estimation. Two literatures in econometrics have highlighted the role for directly estimating Riesz representers: the automatic debiased machine learning literature (as in Chernozhukov et al., 2022b), and an independent literature on sieve methods for conditional moment models (as in Chen et al., 2014). These two literatures solve distinct optimization problems that in the population both have the Riesz representer as their solution. We show that with unregularized or ridge-regularized linear, sieve, or RKHS models, the two resulting estimators are numerically equivalent. However, for other regularization schemes such as the Lasso, or more general machine learning function classes including neural networks, the estimators are not necessarily equivalent. In the latter case, the Chen et al. (2014) formulation yields a novel constrained optimization problem for directly estimating Riesz representers with machine learning. Drawing on results from Birrell et al. (2022), we conjecture that this approach may offer statistical advantages at the cost of greater computational complexity. --- ### MOELIGA: a multi-objective evolutionary approach for feature selection with local improvement Authors: Leandro Vignolo, Matias Gerard Published: 2026-03-21 Categories: cs.NE, cs.LG arXiv: https://arxiv.org/abs/2603.20934v2 Selecting the most relevant or informative features is a key issue in actual machine learning problems. Since an exhaustive search is not feasible even for a moderate number of features, an intelligent search strategy must be employed for finding an optimal subset, which implies considering how features interact with each other in promoting class separability. Balancing feature subset size and classification accuracy constitutes a multi-objective optimization challenge. Here we propose MOELIGA, a multi-objective genetic algorithm incorporating an evolutionary local improvement strategy that evolves subordinate populations to refine feature subsets. MOELIGA employs a crowding-based fitness sharing mechanism and a sigmoid transformation to enhance diversity and guide compactness, alongside a geometry-based objective promoting classifier independence. Experimental evaluation on 14 diverse datasets demonstrates MOELIGA's ability to identify smaller feature subsets with superior or comparable classification performance relative to 11 state-of-the-art methods. These findings suggest MOELIGA effectively addresses the accuracy-dimensionality trade-off, offering a robust and adaptable approach for multi-objective feature selection in complex, high-dimensional scenarios. --- ### AC4A: Access Control for Agents Authors: Reshabh K Sharma, Dan Grossman Published: 2026-03-21 Categories: cs.CR, cs.AI, cs.PL arXiv: https://arxiv.org/abs/2603.20933v1 Large Language Model (LLM) agents combine the chat interaction capabilities of LLMs with the power to interact with external tools and APIs. This enables them to perform complex tasks and act autonomously to achieve user goals. However, current agent systems operate on an all-or-nothing basis: an agent either has full access to an API's capabilities and a web page's content, or it has no access at all. This coarse-grained approach forces users to trust agents with more capabilities than they actually need for a given task. In this paper, we introduce AC4A, an access control framework for agents. As agents become more capable and autonomous, users need a way to limit what APIs or portions of web pages these agents can access, eliminating the need to trust them with everything an API or web page allows. Our goal with AC4A is to provide a framework for defining permissions that lets agents access only the resources they are authorized to access. AC4A works across both API-based and browser-based agents. It does not prescribe what permissions should be, but offers a flexible way to define and enforce them, making it practical for real-world systems. AC4A works by creating permissions granting access to resources, drawing inspiration from established access control frameworks like the one for the Unix file system. Applications define their resources as hierarchies and provide a way to compute the necessary permissions at runtime needed for successful resource access. We demonstrate the usefulness of AC4A in enforcing permissions over real-world APIs and web pages through case studies. The source code of AC4A is available at https://github.com/reSHARMA/AC4A --- ### Causally-Guided Diffusion for Stable Feature Selection Authors: Arun Vignesh Malarkkan, Xinyuan Wang, Kunpeng Liu, Denghui Zhang, Yanjie Fu Published: 2026-03-21 Categories: cs.LG, cs.AI, cs.IT arXiv: https://arxiv.org/abs/2603.20930v1 Feature selection is fundamental to robust data-centric AI, but most existing methods optimize predictive performance under a single data distribution. This often selects spurious features that fail under distribution shifts. Motivated by principles from causal invariance, we study feature selection from a stability perspective and introduce Causally-Guided Diffusion for Stable Feature Selection (CGDFS). In CGDFS, we formalized feature selection as approximate posterior inference over feature subsets, whose posterior mass favors low prediction error and low cross-environment variance. Our framework combines three key insights: First, we formulate feature selection as stability-aware posterior sampling. Here, causal invariance serves as a soft inductive bias rather than explicit causal discovery. Second, we train a diffusion model as a learned prior over plausible continuous selection masks, combined with a stability-aware likelihood that rewards invariance across environments. This diffusion prior captures structural dependencies among features and enables scalable exploration of the combinatorially large selection space. Third, we perform guided annealed Langevin sampling that combines the diffusion prior with the stability objective, which yields a tractable, uncertainty-aware posterior inference that avoids discrete optimization and produces robust feature selections. We evaluate CGDFS on open-source real-world datasets exhibiting distribution shifts. Across both classification and regression tasks, CGDFS consistently selects more stable and transferable feature subsets, which leads to improved out-of-distribution performance and greater selection robustness compared to sparsity-based, tree-based, and stability-selection baselines. --- ### Stability of Sequential and Parallel Coordinate Ascent Variational Inference Authors: Debdeep Pati Published: 2026-03-21 Categories: stat.ML, cs.LG, math.ST, stat.CO arXiv: https://arxiv.org/abs/2603.20929v1 We highlight a striking difference in behavior between two widely used variants of coordinate ascent variational inference: the sequential and parallel algorithms. While such differences were known in the numerical analysis literature in simpler settings, they remain largely unexplored in the optimization-focused literature on variational inference in more complex models. Focusing on the moderately high-dimensional linear regression problem, we show that the sequential algorithm, although typically slower, enjoys convergence guarantees under more relaxed conditions than the parallel variant, which is often employed to facilitate block-wise updates and improve computational efficiency. --- ### Active Inference for Physical AI Agents -- An Engineering Perspective Authors: Bert de Vries Published: 2026-03-21 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2603.20927v1 Physical AI agents, such as robots and other embodied systems operating under tight and fluctuating resource constraints, remain far less capable than biological agents in open-ended real-world environments. This paper argues that Active Inference (AIF), grounded in the Free Energy Principle, offers a principled foundation for closing that gap. We develop this argument from first principles, following a chain from probability theory through Bayesian machine learning and variational inference to active inference and reactive message passing. From the FEP perspective, systems that maintain their structural and functional integrity over time can, under suitable assumptions, be described as minimizing variational free energy (VFE), and AIF operationalizes this by unifying perception, learning, planning, and control within a single computational objective. We show that VFE minimization is naturally realized by reactive message passing on factor graphs, where inference emerges from local, parallel computations. This realization is well matched to the constraints of physical operation, including hard deadlines, asynchronous data, fluctuating power budgets, and changing environments. Because reactive message passing is event-driven, interruptible, and locally adaptable, performance degrades gracefully under reduced resources while model structure can adjust online. We further show that, under suitable coupling and coarse-graining conditions, coupled AIF agents can be described as higher-level AIF agents, yielding a homogeneous architecture based on the same message-passing primitive across scales. Our contribution is not empirical benchmarking, but a clear theoretical and architectural case for the engineering community. --- ### Deep Adaptive Rate Allocation in Volatile Heterogeneous Wireless Networks Authors: Gregorio Maglione, Veselin Rakocevic, Markus Amend, Touraj Soleymani Published: 2026-03-21 Categories: cs.IT, cs.LG arXiv: https://arxiv.org/abs/2603.20926v1 Modern multi-access 5G+ networks provide mobile terminals with additional capacity, improving network stability and performance. However, in highly mobile environments such as vehicular networks, supporting multi-access connectivity remains challenging. The rapid fluctuations of wireless link quality often outpace the responsiveness of existing multipath schedulers and transport-layer protocols. This paper addresses this challenge by integrating Transformer-based path state forecasting with a new multipath splitting scheduler called Deep Adaptive Rate Allocation (DARA). The proposed scheduler employs a deep reinforcement learning engine to dynamically compute optimal congestion window fractions on available paths, determining data allocation among them. A six-component normalised reward function with weight-mediated conflict resolution drives a DQN policy that eliminates the observation-reaction lag inherent in reactive schedulers. Performance evaluation uses a Mininet-based Multipath Datagram Congestion Control Protocol testbed with traces from mobile users in vehicular environments. Experimental results demonstrate that DARA achieves better file transfer time reductions compared to learning-based schedulers under moderate-volatility traces. For buffered video streaming, resolution improvements are maintained across all tested conditions. Under controlled burst scenarios with sub-second buffer constraints, DARA achieves substantial rebuffering improvements whilst state-of-the-art schedulers exhibit near-continuous stalling. --- ### Profit is the Red Team: Stress-Testing Agents in Strategic Economic Interactions Authors: Shouqiao Wang, Marcello Politi, Samuele Marro, Davide Crapis Published: 2026-03-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2603.20925v1 As agentic systems move into real-world deployments, their decisions increasingly depend on external inputs such as retrieved content, tool outputs, and information provided by other actors. When these inputs can be strategically shaped by adversaries, the relevant security risk extends beyond a fixed library of prompt attacks to adaptive strategies that steer agents toward unfavorable outcomes. We propose profit-driven red teaming, a stress-testing protocol that replaces handcrafted attacks with a learned opponent trained to maximize its profit using only scalar outcome feedback. The protocol requires no LLM-as-judge scoring, attack labels, or attack taxonomy, and is designed for structured settings with auditable outcomes. We instantiate it in a lean arena of four canonical economic interactions, which provide a controlled testbed for adaptive exploitability. In controlled experiments, agents that appear strong against static baselines become consistently exploitable under profit-optimized pressure, and the learned opponent discovers probing, anchoring, and deceptive commitments without explicit instruction. We then distill exploit episodes into concise prompt rules for the agent, which make most previously observed failures ineffective and substantially improve target performance. These results suggest that profit-driven red-team data can provide a practical route to improving robustness in structured agent settings with auditable outcomes. --- ### Discriminative Representation Learning for Clinical Prediction Authors: Yang Zhang, Li Fan, Samuel Lawrence, Shi Li Published: 2026-03-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.20921v1 Foundation models in healthcare have largely adopted self supervised pretraining objectives inherited from natural language processing and computer vision, emphasizing reconstruction and large scale representation learning prior to downstream adaptation. We revisit this paradigm in outcome centric clinical prediction settings and argue that, when high quality supervision is available, direct outcome alignment may provide a stronger inductive bias than generative pretraining. We propose a supervised deep learning framework that explicitly shapes representation geometry by maximizing inter class separation relative to within class variance, thereby concentrating model capacity along clinically meaningful axes. Across multiple longitudinal electronic health record tasks, including mortality and readmission prediction, our approach consistently outperforms masked, autoregressive, and contrastive pretraining baselines under matched model capacity. The proposed method improves discrimination, calibration, and sample efficiency, while simplifying the training pipeline to a single stage optimization. These findings suggest that in low entropy, outcome driven healthcare domains, supervision can act as the statistically optimal driver of representation learning, challenging the assumption that large scale self supervised pretraining is a prerequisite for strong clinical performance. --- ### Democratizing AI: A Comparative Study in Deep Learning Efficiency and Future Trends in Computational Processing Authors: Lisan Al Amin, Md Ismail Hossain, Rupak Kumar Das, Mahbubul Islam, Abdulaziz Tabbakh Published: 2026-03-21 Categories: cs.PF, cs.AI, cs.DC, cs.LG arXiv: https://arxiv.org/abs/2603.20920v2 The exponential growth in data has intensified the demand for computational power to train large-scale deep learning models. However, the rapid growth in model size and complexity raises concerns about equal and fair access to computational resources, particularly under increasing energy and infrastructure constraints. GPUs have emerged as essential for accelerating such workloads. This study benchmarks four deep learning models (Conv6, VGG16, ResNet18, CycleGAN) using TensorFlow and PyTorch on Intel Xeon CPUs and NVIDIA Tesla T4 GPUs. Our experiments demonstrate that, on average, GPU training achieves speedups ranging from 11x to 246x depending on model complexity, with lightweight models (Conv6) showing the highest acceleration (246x), mid-sized models (VGG16, ResNet18) achieving 51-116x speedups, and complex generative models (CycleGAN) reaching 11x improvements compared to CPU training. Additionally, in our PyTorch vs. TensorFlow comparison, we observed that TensorFlow's kernel-fusion optimizations reduce inference latency by approximately 15%. We also analyze GPU memory usage trends and projecting requirements through 2025 using polynomial regression. Our findings highlight that while GPUs are essential for sustaining AI's growth, democratized and shared access to GPU resources is critical for enabling research innovation across institutions with limited computational budgets. --- ### Enhancing LIME using Neural Decision Trees Authors: Mohamed Aymen Bouyahia, Argyris Kalogeratos Published: 2026-03-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.20919v1 Interpreting complex machine learning models is a critical challenge, especially for tabular data where model transparency is paramount. Local Interpretable Model-Agnostic Explanations (LIME) has been a very popular framework for interpretable machine learning, also inspiring many extensions. While traditional surrogate models used in LIME variants (e.g. linear regression and decision trees) offer a degree of stability, they can struggle to faithfully capture the complex non-linear decision boundaries that are inherent in many sophisticated black-box models. This work contributes toward bridging the gap between high predictive performance and interpretable decision-making. Specifically, we propose the NDT-LIME variant that integrates Neural Decision Trees (NDTs) as surrogate models. By leveraging the structured, hierarchical nature of NDTs, our approach aims at providing more accurate and meaningful local explanations. We evaluate its effectiveness on several benchmark tabular datasets, showing consistent improvements in explanation fidelity over traditional LIME surrogates. --- ### Do LLM-Driven Agents Exhibit Engagement Mechanisms? Controlled Tests of Information Load, Descriptive Norms, and Popularity Cues Authors: Tai-Quan Peng, Yuan Tian, Songsong Liang, Dazhen Deng, Yingcai Wu Published: 2026-03-21 Categories: cs.AI, cs.CY arXiv: https://arxiv.org/abs/2603.20911v1 Large language models make agent-based simulation more behaviorally expressive, but they also sharpen a basic methodological tension: fluent, human-like output is not, by itself, evidence for theory. We evaluate what an LLM-driven simulation can credibly support using information engagement on social media as a test case. In a Weibo-like environment, we manipulate information load and descriptive norms, while allowing popularity cues (cumulative likes and Sina Weibo-style cumulative reshares) to evolve endogenously. We then ask whether simulated behavior changes in theoretically interpretable ways under these controlled variations, rather than merely producing plausible-looking traces. Engagement responds systematically to information load and descriptive norms, and sensitivity to popularity cues varies across contexts, indicating conditionality rather than rigid prompt compliance. We discuss methodological implications for simulation-based communication research, including multi-condition stress tests, explicit no-norm baselines because default prompts are not blank controls, and design choices that preserve endogenous feedback loops when studying bandwagon dynamics. --- ### LLM-ODE: Data-driven Discovery of Dynamical Systems with Large Language Models Authors: Amirmohammad Ziaei Bideh, Jonathan Gryak Published: 2026-03-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.20910v2 Discovering the governing equations of dynamical systems is a central problem across many scientific disciplines. As experimental data become increasingly available, automated equation discovery methods offer a promising data-driven approach to accelerate scientific discovery. Among these methods, genetic programming (GP) has been widely adopted due to its flexibility and interpretability. However, GP-based approaches often suffer from inefficient exploration of the symbolic search space, leading to slow convergence and suboptimal solutions. To address these limitations, we propose LLM-ODE, a large language model-aided model discovery framework that guides symbolic evolution using patterns extracted from elite candidate equations. By leveraging the generative prior of large language models, LLM-ODE produces more informed search trajectories while preserving the exploratory strengths of evolutionary algorithms. Empirical results on 91 dynamical systems show that LLM-ODE variants consistently outperform classical GP methods in terms of search efficiency and Pareto-front quality. Overall, our results demonstrate that LLM-ODE improves both efficiency and accuracy over traditional GP-based discovery and offers greater scalability to higher-dimensional systems compared to linear and Transformer-only model discovery methods. --- ### Bayesian Scattering: A Principled Baseline for Uncertainty on Image Data Authors: Bernardo Fichera, Zarko Ivkovic, Kjell Jorner, Philipp Hennig, Viacheslav Borovitskiy Published: 2026-03-21 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2603.20908v1 Uncertainty quantification for image data is dominated by complex deep learning methods, yet the field lacks an interpretable, mathematically grounded baseline. We propose Bayesian scattering to fill this gap, serving as a first-step baseline akin to the role of Bayesian linear regression for tabular data. Our method couples the wavelet scattering transform-a deep, non-learned feature extractor-with a simple probabilistic head. Because scattering features are derived from geometric principles rather than learned, they avoid overfitting the training distribution. This helps provide sensible uncertainty estimates even under significant distribution shifts. We validate this on diverse tasks, including medical imaging under institution shift, wealth mapping under country-to-country shift, and Bayesian optimization of molecular properties. Our results suggest that Bayesian scattering is a solid baseline for complex uncertainty quantification methods. --- ### The Hidden Puppet Master: Predicting Human Belief Change in Manipulative LLM Dialogues Authors: Jocelyn Shen, Amina Luvsanchultem, Jessica Kim, Kynnedy Smith, Valdemar Danry, Kantwon Rogers, Hae Won Park, Maarten Sap, Cynthia Breazeal Published: 2026-03-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.20907v5 As users increasingly turn to LLMs for practical and personal advice, they become vulnerable to subtle steering toward hidden incentives misaligned with their own interests. While existing NLP research has benchmarked manipulation detection, these efforts often rely on simulated debates and remain fundamentally decoupled from actual human belief shifts in real-world scenarios. We introduce PUPPET, a theoretical taxonomy and resource that bridges this gap by focusing on the moral direction of hidden incentives in everyday, advice-giving contexts. We provide an evaluation dataset of N=1,035 human-LLM interactions, where we measure users' belief shifts. Our analysis reveals a critical disconnect in current safety paradigms: while models can be trained to detect manipulative strategies, they do not correlate with the magnitude of resulting belief change. As such, we define the task of human belief shift prediction and show that while state-of-the-art LLMs achieve moderate correlation (r=0.3-0.5), they exhibit systematic directional biases, with certain models over or under-predicting the magnitude of human belief change. This work establishes a theoretically grounded and behaviorally validated foundation for AI social safety efforts by studying incentive-driven manipulation in LLMs during everyday, practical user queries. --- ### The Hidden Puppet Master: Predicting Human Belief Change in Manipulative LLM Dialogues Authors: Jocelyn Shen, Amina Luvsanchultem, Jessica Kim, Kynnedy Smith, Valdemar Danry, Kantwon Rogers, Hae Won Park, Maarten Sap, Cynthia Breazeal Published: 2026-03-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.20907v4 As users increasingly turn to LLMs for practical and personal advice, they become vulnerable to subtle steering toward hidden incentives misaligned with their own interests. While existing NLP research has benchmarked manipulation detection, these efforts often rely on simulated debates and remain fundamentally decoupled from actual human belief shifts in real-world scenarios. We introduce PUPPET, a theoretical taxonomy and resource that bridges this gap by focusing on the moral direction of hidden incentives in everyday, advice-giving contexts. We provide an evaluation dataset of N=1,035 human-LLM interactions, where we measure users' belief shifts. Our analysis reveals a critical disconnect in current safety paradigms: while models can be trained to detect manipulative strategies, they do not correlate with the magnitude of resulting belief change. As such, we define the task of human belief shift prediction and show that while state-of-the-art LLMs achieve moderate correlation (r=0.3-0.5), they exhibit systematic directional biases, with certain models over or under-predicting the magnitude of human belief change. This work establishes a theoretically grounded and behaviorally validated foundation for AI social safety efforts by studying incentive-driven manipulation in LLMs during everyday, practical user queries. --- ### Data-Driven Weak-form Discovery of Stochastic Systems Authors: Eshwar R A, Gajanan V. Honnavar Published: 2026-03-21 Categories: stat.ME, math-ph, math.DS, nlin.CD, physics.data-an, stat.ML arXiv: https://arxiv.org/abs/2603.20904v6 We present an algorithm for learning the governing equations of a stochastic dynamical system from trajectory data. It recovers interpretable symbolic expressions for both the drift $b(x)$ and the diffusion $a(x)$ in a single pass, yielding a model that can be queried directly for relaxation timescales, metastable escape rates, and stationary distributions. Rather than estimating the dynamics one time step at a time, the algorithm averages each candidate term across the whole trajectory before regressing; a drift-informed correction further removes the finite-sampling bias in the diffusion estimate, cutting it from 4.6% to 0.6% for state-dependent noise. We also show that the trajectory averaging must use a spatial rather than a temporal weighting: temporal weighting, as in existing weak-form methods, is biased for stochastic data with an error that grows with dataset size. On three benchmark systems -- the Ornstein--Uhlenbeck process, a double-well Langevin system, and a multiplicative-noise system -- the algorithm recovers all coefficients to within 5%, stationary densities to within 0.01 in total variation, and escape rates that match the true dynamics. --- ### Unfolding with a Wasserstein Loss Authors: Katy Craig, Benjamin Faktor, Benjamin Nachman Published: 2026-03-21 Categories: math.OC, hep-ph, stat.ML arXiv: https://arxiv.org/abs/2603.20903v2 Data unfolding -- the removal of noise or artifacts from measurements -- is a fundamental task across the experimental sciences. Of particular interest are applications in physics, where the dominant approach is Richardson-Lucy (RL) deconvolution. The classical RL approach aims to find denoised data that, once passed through the noise model, is as close as possible to the measured data in terms of Kullback-Leibler (KL) divergence. This requires that the support of the measured data overlaps with the output of the noise model, a hypothesis typically enforced by binning, which introduces numerical error. As a counterpoint, the present work studies an alternative formulation using a Wasserstein loss. We establish sharp conditions for existence and uniqueness of optimizers, answering open questions of Li, et al., regarding necessary conditions for uniqueness in the case of transport map noise models. We then develop a provably convergent generalized Sinkhorn algorithm to compute approximate optimizers. Our algorithm requires only empirical observations of the noise model and measured data and scales with the size of the data, rather than the ambient dimension. Numerical experiments on one- and two-dimensional problems inspired by jet mass unfolding in particle physics demonstrate that the optimal transport approach offers robust, accurate performance compared to classical RL deconvolution, particularly when binning artifacts are significant. --- ### Mitigating Shortcut Reasoning in Language Models: A Gradient-Aware Training Approach Authors: Hongyu Cao, Kunpeng Liu, Dongjie Wang, Yanjie Fu Published: 2026-03-21 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.20899v1 Large language models exhibit strong reasoning capabilities, yet often rely on shortcuts such as surface pattern matching and answer memorization rather than genuine logical inference. We propose Shortcut-Aware Reasoning Training (SART), a gradient-aware framework that detects and mitigates shortcut-promoting samples via ShortcutScore and gradient surgery. Our method identifies shortcut signals through gradient misalignment with validation objectives and answer-token concentration, and modifies training dynamics accordingly. Experiments on controlled reasoning benchmarks show that SART achieves +16.5% accuracy and +40.2% robustness over the strongest baseline, significantly improving generalization under distribution shifts. Code is available at: https://github.com/fuyanjie/short-cut-aware-data-centric-reasoning. --- ### Natural Gradient Descent for Online Continual Learning Authors: Joe Khawand, David Colliaux Published: 2026-03-21 Categories: cs.LG, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2603.20898v1 Online Continual Learning (OCL) for image classification represents a challenging subset of Continual Learning, focusing on classifying images from a stream without assuming data independence and identical distribution (i.i.d). The primary challenge in this context is to prevent catastrophic forgetting, where the model's performance on previous tasks deteriorates as it learns new ones. Although various strategies have been proposed to address this issue, achieving rapid convergence remains a significant challenge in the online setting. In this work, we introduce a novel approach to training OCL models that utilizes the Natural Gradient Descent optimizer, incorporating an approximation of the Fisher Information Matrix (FIM) through Kronecker Factored Approximate Curvature (KFAC). This method demonstrates substantial improvements in performance across all OCL methods, particularly when combined with existing OCL tricks, on datasets such as Split CIFAR-100, CORE50, and Split miniImageNet. --- ### The data heat island effect: quantifying the impact of AI data centers in a warming world Authors: Andrea Marinoni, Erik Cambria, Weisi Lin, Mauro Dalla Mura, Jocelyn Chanussot, Edoardo Ragusa, Chi Yan Tso, Yihao Zhu, Benjamin Horton Published: 2026-03-21 Categories: cs.CY, cs.AI, cs.AR arXiv: https://arxiv.org/abs/2603.20897v3 The strong and continuous increase of AI-based services leads to the steady proliferation of AI data centres worldwide with the unavoidable escalation of their power consumption. It is unknown how this energy demand for computational purposes will impact the surrounding environment. Here, we focus our attention on the heat dissipation of AI hyperscalers. Taking advantage of land surface temperature measurements acquired by remote sensing platforms over the last decades, we are able to obtain a robust assessment of the temperature increase recorded in the areas surrounding AI data centres globally. We estimate that the land surface temperature increases by 2°C on average after the start of operations of an AI data centre, inducing local microclimate zones, which we call the data heat island effect. We assess the impact on the communities, quantifying that more than 340 million people could be affected by this temperature increase. Our results show that the data heat island effect could have a remarkable influence on communities and regional welfare in the future, hence becoming part of the conversation around environmentally sustainable AI worldwide. --- ### Spectral-Sphere-Constrained Hyper-Connections Authors: Zhaoyi Liu, Haichuan Zhang, Ang Li Published: 2026-03-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.20896v2 Hyper-Connections (HC) extend residual connections into multiple streams, employing residual matrices for cross-stream mixing to enrich model expressivity. However, unconstrained mixing disrupts the identity mapping property intrinsic to the residual connection, causing unstable training. To address this, Manifold-Constrained Hyper-Connections (mHC) and its variants restrict these matrices to be doubly stochastic via Sinkhorn-Knopp (SK) algorithm or permutation-based parameterizations. We reveal three limitations of this doubly stochastic constraint: (1) identity degeneration, where learned matrices collapse around the identity initialization and diminish cross-stream interactions, (2) a expressivity bottleneck, where the doubly stochastic constraint restricts the freedom of the subdominant spectrum of the residual matrices, preventing the model from selectively preserving or attenuating cross-stream variations, and (3) parameterization inefficiencies, manifesting as unstable SK iterations or the factorial-scaling overhead of permutation-based parameterizations. To overcome these flaws, we propose Spectral-Sphere-Constrained Hyper-Connections ($\mathrm{s}^{2}$HC). By confining residual matrices to a spectral norm sphere, $\mathrm{s}^{2}$HC restores free control over the subdominant spectrum, enabling the model to selectively preserve or attenuate cross-stream variations. This shift eliminates unstable SK iterations and factorial parameterization, enabling expressive, non-degenerate residual matrices while preserving training stability. --- ### Beyond the Birkhoff Polytope: Spectral-Sphere-Constrained Hyper-Connections Authors: Zhaoyi Liu, Haichuan Zhang, Ang Li Published: 2026-03-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.20896v1 Hyper-Connections (HC) generalize residual connections into multiple streams, employing residual matrices for cross-stream feature mixing to enrich model expressivity. However, unconstrained mixing disrupts the identity mapping property intrinsic to the residual connection, causing unstable training. To address this, Manifold-Constrained Hyper-Connections (mHC) and its variant restrict these matrices to the Birkhoff polytope (doubly stochastic matrices) via Sinkhorn iterations or permutation-based parameterizations. We reveal three limitations of this polytope constraint: (1) identity degeneration, where learned matrices collapse around the identity and diminish cross-stream interactions, (2) an expressivity bottleneck, as the non-negativity constraint prevents subtractive feature disentanglement, and (3) parameterization inefficiencies, manifesting as unstable Sinkhorn iterations or the factorial-scaling overhead of permutation-based parameterizations. To overcome these flaws, we propose Spectral-Sphere-Constrained Hyper-Connections (sHC). By geometrically shifting the feasible set from a rigid polytope to a spectral norm sphere, sHC allows negative entries, unlocking subtractive interactions for selective feature diversification. This shift eliminates unstable Sinkhorn projections and factorial parameterization, enabling expressive, non-degenerate residual matrices while preserving training stability. --- ### LLM Router: Rethinking Routing with Prefill Activations Authors: Tanay Varshney, Annie Surla, Michelle Xu, Gomathy Venkata Krishnan, Maximilian Jeblick, David Austin, Neal Vaidya, Davide Onofrio Published: 2026-03-21 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2603.20895v3 Existing routers rely on semantic query features or handcrafted features, which often fail to capture model-specific failures or intrinsic task difficulty. We instead route using internal LLM activations, specifically the residual stream. Our key idea, Encoder-Target Decoupling, separates the model that produces the predictive signal (the Encoder) from the model whose correctness is being estimated (the Target), allowing open-weight encoders to predict the performance of closed-source target models. We evaluate layerwise geometric probes, finding that Fisher Separability ($J$) effectively identifies informative layers, supported by Effective Dimensionality ($d_{\mathrm{eff}}$) diagnostics. We then utilize a SharedTrunkNet, a joint multi-output MLP that predicts simultaneous correctness probabilities across candidate models using concatenated prefill features. In our experiments, SharedTrunkNet consistently outperforms semantic baselines. At its best, SharedTrunkNet closes 45.58% of the gap between the strongest standalone model and the oracle while achieving 74.31% cost savings relative to the most expensive model. These results demonstrate that prefill activations provide a robust routing signal, establishing activation-based routing as a high-performance alternative to purely semantic selection. --- ### LLM Router: Rethinking Routing with Prefill Activations Authors: Tanay Varshney, Annie Surla, Michelle Xu, Gomathy Venkata Krishnan, Maximilian Jeblick, David Austin, Neal Vaidya, Davide Onofrio Published: 2026-03-21 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2603.20895v2 LLMs often achieve similar average benchmark accuracies while exhibiting complementary strengths on different subsets of queries, suggesting that a router with query-specific model selection can outperform any single model. While existing routers rely on semantic query features, they often fail to capture model-specific failures or intrinsic task difficulty. We instead study routing via internal prefill activations. Our key idea, Encoder-Target Decoupling, separates the model that produces the predictive signal (the Encoder) from the model whose correctness is being estimated (the Target), allowing open-weight encoders to predict the performance of closed-source target models. We evaluate layerwise geometric probes, finding that Fisher Separability (J) effectively identifies informative layers, supported by Effective Dimensionality (d_eff) diagnostics. We then utilize a SharedTrunkNet, a joint multi-output MLP that predicts simultaneous correctness probabilities across candidate models using concatenated prefill features. In our experiments, SharedTrunkNet consistently outperforms semantic baselines. At its best, SharedTrunkNet closes 45.58% of the gap between the strongest standalone model and the oracle while achieving 74.31% cost savings relative to the most expensive model. These results demonstrate that prefill activations provide a robust routing signal, establishing mechanistic routing as a high-performance alternative to purely semantic selection. --- ### Auto-differentiable data assimilation: Co-learning of states, dynamics, and filtering algorithms Authors: Melissa Adrian, Daniel Sanz-Alonso, Rebecca Willett Published: 2026-03-21 Categories: stat.ML, cs.LG, eess.SP, math.DS arXiv: https://arxiv.org/abs/2603.20891v1 Data assimilation algorithms estimate the state of a dynamical system from partial observations, where the successful performance of these algorithms hinges on costly parameter tuning and on employing an accurate model for the dynamics. This paper introduces a framework for jointly learning the state, dynamics, and parameters of filtering algorithms in data assimilation through a process we refer to as auto-differentiable filtering. The framework leverages a theoretically motivated loss function that enables learning from partial, noisy observations via gradient-based optimization using auto-differentiation. We further demonstrate how several well-known data assimilation methods can be learned or tuned within this framework. To underscore the versatility of auto-differentiable filtering, we perform experiments on dynamical systems spanning multiple scientific domains, such as the Clohessy-Wiltshire equations from aerospace engineering, the Lorenz-96 system from atmospheric science, and the generalized Lotka-Volterra equations from systems biology. Finally, we provide guidelines for practitioners to customize our framework according to their observation model, accuracy requirements, and computational budget. --- ### Scene Graph-guided SegCaptioning Transformer with Fine-grained Alignment for Controllable Video Segmentation and Captioning Authors: Xu Zhang, Jin Yuan, BinHong Yang, Xuan Liu, Qianjun Zhang, Yuyi Wang, Zhiyong Li, Hanwang Zhang Published: 2026-03-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.20887v1 Recent advancements in multimodal large models have significantly bridged the representation gap between diverse modalities, catalyzing the evolution of video multimodal interpretation, which enhances users' understanding of video content by generating correlated modalities. However, most existing video multimodal interpretation methods primarily concentrate on global comprehension with limited user interaction. To address this, we propose a novel task, Controllable Video Segmentation and Captioning (SegCaptioning), which empowers users to provide specific prompts, such as a bounding box around an object of interest, to simultaneously generate correlated masks and captions that precisely embody user intent. An innovative framework Scene Graph-guided Fine-grained SegCaptioning Transformer (SG-FSCFormer) is designed that integrates a Prompt-guided Temporal Graph Former to effectively captures and represents user intent through an adaptive prompt adaptor, ensuring that the generated content well aligns with the user's requirements. Furthermore, our model introduces a Fine-grained Mask-linguistic Decoder to collaboratively predict high-quality caption-mask pairs using a Multi-entity Contrastive loss, as well as provide fine-grained alignment between each mask and its corresponding caption tokens, thereby enhancing users' comprehension of videos. Comprehensive experiments conducted on two benchmark datasets demonstrate that SG-FSCFormer achieves remarkable performance, effectively capturing user intent and generating precise multimodal outputs tailored to user specifications. Our code is available at https://github.com/XuZhang1211/SG-FSCFormer. --- ### Characterizing the onset and offset of motor imagery during passive arm movements induced by an upper-body exoskeleton Authors: Kanishka Mitra, Frigyes Samuel Racz, Satyam Kumar, Ashish D. Deshpande, José del R. Millán Published: 2026-03-21 Categories: cs.RO, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2603.20885v1 Two distinct technologies have gained attention lately due to their prospects for motor rehabilitation: robotics and brain-machine interfaces (BMIs). Harnessing their combined efforts is a largely uncharted and promising direction that has immense clinical potential. However, a significant challenge is whether motor intentions from the user can be accurately detected using non-invasive BMIs in the presence of instrumental noise and passive movements induced by the rehabilitation exoskeleton. As an alternative to the straightforward continuous control approach, this study instead aims to characterize the onset and offset of motor imagery during passive arm movements induced by an upper-body exoskeleton to allow for the natural control (initiation and termination) of functional movements. Ten participants were recruited to perform kinesthetic motor imagery (MI) of the right arm while attached to the robot, simultaneously cued with LEDs indicating the initiation and termination of a goal-oriented reaching task. Using electroencephalogram signals, we built a decoder to detect the transition between i) rest and beginning MI and ii) maintaining and ending MI. Offline decoder evaluation achieved group average onset accuracy of 60.7% and 66.6% for offset accuracy, revealing that the start and stop of MI could be identified while attached to the robot. Furthermore, pseudo-online evaluation could replicate this performance, forecasting reliable online exoskeleton control in the future. Our approach showed that participants could produce quality and reliable sensorimotor rhythms regardless of noise or passive arm movements induced by wearing the exoskeleton, which opens new possibilities for BMI control of assistive devices. --- ### MemoNoveltyAgent: A Historical Research Memory-Aware Agent Workflow for Paper Novelty Assessment Authors: Jiajun Hou, Hexuan Deng, Wenxiang Jiao, Xuebo Liu, Xiaopeng Ke, Derek F. Wong, Min Zhang Published: 2026-03-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.20884v4 To alleviate the heavy burden of paper screening, researchers increasingly rely on existing AI agents, such as AI reviewers or DeepResearch, for paper evaluation and novelty assessment. However, lacking specialized mechanisms for processing scholarly literature, their analyses often produce superficial results with noticeable deficiencies in quality. To bridge this gap, we introduce MemoNoveltyAgent, a multi-agent system designed to generate comprehensive and faithful novelty reports. Beyond retrieving concrete prior-paper evidence via RAG, our system incorporates a high-level abstract memory constructed from large-scale scholarly corpora. This memory organizes research into hierarchical trees to distill field-specific evolutionary trajectories, thereby providing a broader historical context. Furthermore, we decompose papers into discrete novelty points for fine-grained analysis and retrieval, while employing a self-validation mechanism to improve report faithfulness. Finally, to address the evaluation challenges of such open-ended generation tasks, we propose a RAG-augmented checklist evaluation method that enables reliable and evidence-grounded assessments. Extensive experiments demonstrate that MemoNoveltyAgent outperforms GPT-5 DeepResearch by 13.69%. Code and demo are available at https://github.com/SStan1/MemoNoveltyAgent --- ### MemoNoveltyAgent: A Historical Research Memory-Aware Agent Workflow for Paper Novelty Assessment Authors: Jiajun Hou, Hexuan Deng, Wenxiang Jiao, Xuebo Liu, Xiaopeng Ke, Derek F. Wong, Min Zhang Published: 2026-03-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.20884v3 To alleviate the heavy burden of paper screening, researchers increasingly rely on existing AI agents, such as AI reviewers or DeepResearch, for paper evaluation and novelty assessment. However, lacking specialized mechanisms for processing scholarly literature, their analyses often produce superficial results with noticeable deficiencies in quality. To bridge this gap, we introduce MemoNoveltyAgent, a multi-agent system designed to generate comprehensive and faithful novelty reports. Beyond retrieving concrete prior-paper evidence via RAG, our system incorporates a high-level abstract memory constructed from large-scale scholarly corpora. This memory organizes research into hierarchical trees to distill field-specific evolutionary trajectories, thereby providing a broader historical context. Furthermore, we decompose papers into discrete novelty points for fine-grained analysis and retrieval, while employing a self-validation mechanism to improve report faithfulness. Finally, to address the evaluation challenges of such open-ended generation tasks, we propose a RAG-augmented checklist evaluation method that enables reliable and evidence-grounded assessments. Extensive experiments demonstrate that MemoNoveltyAgent outperforms GPT-5 DeepResearch by 13.69%. Code and demo are available at https://github.com/SStan1/MemoNoveltyAgent --- ### RubricRAG: Towards Interpretable and Reliable LLM Evaluation via Domain Knowledge Retrieval for Rubric Generation Authors: Kaustubh D. Dhole, Eugene Agichtein Published: 2026-03-21 Categories: cs.IR, cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2603.20882v1 Large language models (LLMs) are increasingly evaluated and sometimes trained using automated graders such as LLM-as-judges that output scalar scores or preferences. While convenient, these approaches are often opaque: a single score rarely explains why an answer is good or bad, which requirements were missed, or how a system should be improved. This lack of interpretability limits their usefulness for model development, dataset curation, and high-stakes deployment. Query-specific rubric-based evaluation offers a more transparent alternative by decomposing quality into explicit, checkable criteria. However, manually designing high-quality, query-specific rubrics is labor-intensive and cognitively demanding and not feasible for deployment. While previous approaches have focused on generating intermediate rubrics for automated downstream evaluation, it is unclear if these rubrics are both interpretable and effective for human users. In this work, we investigate whether LLMs can generate useful, instance-specific rubrics as compared to human-authored rubrics, while also improving effectiveness for identifying good responses. Through our systematic study on two rubric benchmarks, and on multiple few-shot and post-training strategies, we find that off-the-shelf LLMs produce rubrics that are poorly aligned with human-authored ones. We introduce a simple strategy, RubricRAG, which retrieves domain knowledge via rubrics at inference time from related queries. We demonstrate that RubricRAG can generate more interpretable rubrics both for similarity to human-authored rubrics, and for improved downstream evaluation effectiveness. Our results highlight both the challenges and a promising approach of scalable, interpretable evaluation through automated rubric generation. --- ### Incentive-Aware Federated Averaging with Performance Guarantees under Strategic Participation Authors: Fateme Maleki, Krishnan Raghavan, Farzad Yousefian Published: 2026-03-21 Categories: cs.LG, math.OC arXiv: https://arxiv.org/abs/2603.20873v2 Federated learning (FL) is a communication-efficient collaborative learning framework that enables model training across multiple agents with private local datasets. While the benefits of FL in improving global model performance are well established, individual agents may behave strategically, balancing the learning payoff against the cost of contributing their local data. Motivated by the need for FL frameworks that successfully retain participating agents, we propose an incentive-aware federated averaging method in which, at each communication round, clients transmit both their local model parameters and their updated training dataset sizes to the server. The dataset sizes are dynamically adjusted via a Nash equilibrium (NE)-seeking update rule that captures strategic data participation. We analyze the proposed method under convex and nonconvex global objective settings and establish performance guarantees for the resulting incentive-aware FL algorithm. Furthermore, under a merely monotone game setting, we consider a welfare loss minimization framework and establish asymptotic convergence of the scheme. Numerical experiments on the MNIST and CIFAR-10 datasets demonstrate that agents achieve competitive global model performance while converging to stable data participation strategies. --- ### ReLaMix: Residual Latency-Aware Mixing for Delay-Robust Financial Time-Series Forecasting Authors: Tianyou Lai, Wentao Yue, Jiayi Zhou, Chaoyuan Hao, Lingke Chang, Qingyu Mao, Zhibo Niu, Qilei Li Published: 2026-03-21 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2603.20869v1 Financial time-series forecasting in real-world high-frequency markets is often hindered by delayed or partially stale observations caused by asynchronous data acquisition and transmission latency. To better reflect such practical conditions, we investigate a simulated delay setting where a portion of historical signals is corrupted by a Zero-Order Hold (ZOH) mechanism, significantly increasing forecasting difficulty through stepwise stagnation artifacts. In this paper, we propose ReLaMix (Residual Latency-Aware Mixing Network), a lightweight extension of TimeMixer that integrates learnable bottleneck compression with residual refinement for robust signal recovery under delayed observations. ReLaMix explicitly suppresses redundancy from repeated stale values while preserving informative market dynamics via residual mixing enhancement. Experiments on a large-scale second-resolution PAXGUSDT benchmark demonstrate that ReLaMix consistently achieves state-of-the-art accuracy across multiple delay ratios and prediction horizons, outperforming strong mixer and Transformer baselines with substantially fewer parameters. Moreover, additional evaluations on BTCUSDT confirm the cross-asset generalization ability of the proposed framework. These results highlight the effectiveness of residual bottleneck mixing for high-frequency financial forecasting under realistic latency-induced staleness. --- ### TAFG-MAN: Timestep-Adaptive Frequency-Gated Latent Diffusion for Efficient and High-Quality Low-Dose CT Image Denoising Authors: Tangtangfang Fang, Yang Jiao, Xiangjian He, Jingxi Hu, Jiaqi Yang Published: 2026-03-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.20868v1 Low-dose computed tomography (LDCT) reduces radiation exposure but also introduces substantial noise and structural degradation, making it difficult to suppress noise without erasing subtle anatomical details. In this paper, we present TAFG-MAN, a latent diffusion framework for efficient and high-quality LDCT image denoising. The framework combines a perceptually optimized autoencoder, conditional latent diffusion restoration in a compact latent space, and a lightweight Timestep-Adaptive Frequency-Gated (TAFG) conditioning design. TAFG decomposes condition features into low- and high-frequency components, predicts timestep-adaptive gates from the current denoising feature and timestep embedding, and progressively releases high-frequency guidance in later denoising stages before cross-attention. In this way, the model relies more on stable structural guidance at early reverse steps and introduces fine details more cautiously as denoising proceeds, improving the balance between noise suppression and detail preservation. Experiments show that TAFG-MAN achieves a favorable quality-efficiency trade-off against representative baselines. Compared with its base variant without TAFG, it further improves detail preservation and perceptual quality while maintaining essentially the same inference cost, and ablation results confirm the effectiveness of the proposed conditioning mechanism. --- ### Semantic Sections: An Atlas-Native Feature Ontology for Obstructed Representation Spaces Authors: Hossein Javidnia Published: 2026-03-21 Categories: cs.LG, cs.AI, cs.CL, cs.NE arXiv: https://arxiv.org/abs/2603.20867v1 Recent interpretability work often treats a feature as a single global direction, dictionary atom, or latent coordinate shared across contexts. We argue that this ontology can fail in obstructed representation spaces, where locally coherent meanings need not assemble into one globally consistent feature. We introduce an atlas-native replacement object, the semantic section: a transport-compatible family of local feature representatives defined over a context atlas. We formalize semantic sections, prove that tree-supported propagation is always pathwise realizable, and show that cycle consistency is the key criterion for genuine globalization. This yields a distinction between tree-local, globalizable, and twisted sections, with twisted sections capturing locally coherent but holonomy-obstructed meanings. We then develop a discovery-and-certification pipeline based on seeded propagation, synchronization across overlaps, defect-based pruning, cycle-aware taxonomy, and deduplication. Across layer-16 atlases for Llama 3.2 3B Instruct, Qwen 2.5 3B Instruct, and Gemma 2 2B IT, we find nontrivial populations of semantic sections, including cycle-supported globalizable and twisted regimes after deduplication. Most importantly, semantic identity is not recovered by raw global-vector similarity. Even certified globalizable sections show low cross-chart signed cosine similarity, and raw similarity baselines recover only a small fraction of true within-section pairs, often collapsing at moderate thresholds. By contrast, section-based identity recovery is perfect on certified supports. These results support semantic sections as a better feature ontology in obstructed regimes. --- ### Discrete Coefficients and Open Invariant Covers in the Homology of Ample Groupoids Authors: Luciano Melodia Published: 2026-03-21 Categories: math.AT, cs.LG, math.KT, math.OA arXiv: https://arxiv.org/abs/2603.20861v2 The homology of an ample groupoid is computed from the complex of compactly supported continuous functions on the nerve. Two hypotheses routinely imposed on this complex behave in opposite ways. We show that the comparison map from the integral chain complex tensored with the coefficient group to the chain complex with coefficients is always injective, and that it is surjective if and only if every compactly supported continuous function into the coefficient group is locally constant. The universal coefficient theorem for ample groupoids is therefore a discrete coefficient statement, and it already fails for the real numbers on the Cantor set. We then show that a cover of the unit space by two clopen invariant subsets forces the groupoid to split as a disjoint union of three clopen invariant reductions, so that the associated Mayer-Vietoris sequence has vanishing connecting maps and computes nothing. Dropping closedness repairs this, since two open invariant subsets covering the unit space give a Mayer-Vietoris sequence for every topological abelian coefficient group. For an integer action on a two point compactification of the integers we compute that sequence and find that its connecting map is an isomorphism of infinite cyclic groups, which no cover of a unit space by clopen invariant subsets can achieve. --- ### Universal Coefficients and Mayer-Vietoris for Moore Homology of Ample Groupoids Authors: Luciano Melodia Published: 2026-03-21 Categories: math.AT, cs.LG, math.KT, math.OA arXiv: https://arxiv.org/abs/2603.20861v1 We establish two structural results for Moore homology of ample groupoids. First, for every ample groupoid $\mathcal{G}$ and every discrete abelian coefficient group $A$, we prove a universal coefficient theorem relating the homology groups $H_n(\mathcal{G};A)$ to the integral Moore homology of $\mathcal{G}$. More precisely, we obtain a natural short exact sequence $$ 0 \longrightarrow H_n(\mathcal{G};\mathbb{Z})\otimes_{\mathbb{Z}} A \xrightarrow{κ_n^{\mathcal{G}}} H_n(\mathcal{G};A) \xrightarrow{ι_n^{\mathcal{G}}} \operatorname{Tor}_1^{\mathbb{Z}}\bigl(H_{n-1}(\mathcal{G};\mathbb{Z}),A\bigr) \longrightarrow 0. $$ Second, for a decomposition of the unit space into clopen saturated subsets, we prove a Mayer-Vietoris long exact sequence in Moore homology. The proof is carried out at the chain level and is based on a short exact sequence of Moore chain complexes associated to the corresponding restricted groupoids. These results provide effective tools for the computation of Moore homology. We also explain why the discreteness of the coefficient group is essential for the universal coefficient theorem. --- ### Restoring Neural Network Plasticity for Faster Transfer Learning Authors: Xander Coetzer, Arné Schreuder, Anna Sergeevna Bosman Published: 2026-03-21 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.20860v1 Transfer learning with models pretrained on ImageNet has become a standard practice in computer vision. Transfer learning refers to fine-tuning pretrained weights of a neural network on a downstream task, typically unrelated to ImageNet. However, pretrained weights can become saturated and may yield insignificant gradients, failing to adapt to the downstream task. This hinders the ability of the model to train effectively, and is commonly referred to as loss of neural plasticity. Loss of plasticity may prevent the model from fully adapting to the target domain, especially when the downstream dataset is atypical in nature. While this issue has been widely explored in continual learning, it remains relatively understudied in the context of transfer learning. In this work, we propose the use of a targeted weight re-initialization strategy to restore neural plasticity prior to fine-tuning. Our experiments show that both convolutional neural networks (CNNs) and vision transformers (ViTs) benefit from this approach, yielding higher test accuracy with faster convergence on several image classification benchmarks. Our method introduces negligible computational overhead and is compatible with common transfer learning pipelines. --- ### Fast and Robust Deformable 3D Gaussian Splatting Authors: Han Jiao, Jiakai Sun, Lei Zhao, Zhanjie Zhang, Wei Xing, Huaizhong Lin Published: 2026-03-21 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2603.20857v1 3D Gaussian Splatting has demonstrated remarkable real-time rendering capabilities and superior visual quality in novel view synthesis for static scenes. Building upon these advantages, researchers have progressively extended 3D Gaussians to dynamic scene reconstruction. Deformation field-based methods have emerged as a promising approach among various techniques. These methods maintain 3D Gaussian attributes in a canonical field and employ the deformation field to transform this field across temporal sequences. Nevertheless, these approaches frequently encounter challenges such as suboptimal rendering speeds, significant dependence on initial point clouds, and vulnerability to local optima in dim scenes. To overcome these limitations, we present FRoG, an efficient and robust framework for high-quality dynamic scene reconstruction. FRoG integrates per-Gaussian embedding with a coarse-to-fine temporal embedding strategy, accelerating rendering through the early fusion of temporal embeddings. Moreover, to enhance robustness against sparse initializations, we introduce a novel depth- and error-guided sampling strategy. This strategy populates the canonical field with new 3D Gaussians at low-deviation initial positions, significantly reducing the optimization burden on the deformation field and improving detail reconstruction in both static and dynamic regions. Furthermore, by modulating opacity variations, we mitigate the local optima problem in dim scenes, improving color fidelity. Comprehensive experimental results validate that our method achieves accelerated rendering speeds while maintaining state-of-the-art visual quality. --- ### Ensemble of Small Classifiers For Imbalanced White Blood Cell Classification Authors: Siddharth Srivastava, Adam Smith, Scott Brooks, Jack Bacon, Till Bretschneider Published: 2026-03-21 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2603.20856v1 Automating white blood cell classification for diagnosis of leukaemia is a promising alternative to time-consuming and resource-intensive examination of cells by expert pathologists. However, designing robust algorithms for classification of rare cell types remains challenging due to variations in staining, scanning and inter-patient heterogeneity. We propose a lightweight ensemble approach for classification of cells during Haematopoiesis, with a focus on the biology of Granulopoiesis, Monocytopoiesis and Lymphopoiesis. Through dataset expansion to alleviate some class imbalance, we demonstrate that a simple ensemble of lightweight pretrained SwinV2-Tiny, DinoBloom-Small and ConvNeXT-V2-Tiny models achieves excellent performance on this challenging dataset. We train 3 instantiations of each architecture in a stratified 3-fold cross-validation framework; for an input image, we forward-pass through all 9 models and aggregate through logit averaging. We further reason on the weaknesses of our model in confusing similar-looking myelocytes in granulopoiesis and lymphocytes in lymphopoiesis. Code: https://gitlab.com/siddharthsrivastava/wbc-bench-2026. --- ### SozKZ: Training Efficient Small Language Models for Kazakh from Scratch Authors: Saken Tukenov Published: 2026-03-21 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.20854v1 Kazakh, a Turkic language spoken by over 22 million people, remains underserved by existing multilingual language models, which allocate minimal capacity to low-resource languages and employ tokenizers ill-suited to agglutinative morphology. We present SozKZ, a family of Llama-architecture language models (50M-600M parameters) trained entirely from scratch on 9 billion tokens of Kazakh text with a dedicated 50K BPE tokenizer. We evaluate all models on three Kazakh benchmarks -- multiple-choice cultural QA, reading comprehension (Belebele), and topic classification (SIB-200) -- alongside five multilingual baselines ranging from 500M to 3B parameters. Our 600M model achieves 30.3% accuracy on Kazakh cultural QA, approaching the 32.0% of Llama-3.2-1B (2x larger), and 25.5% on SIB-200 topic classification, surpassing all evaluated multilingual models up to 2B parameters. We observe consistent scaling from 50M to 600M, with MC QA accuracy rising from 22.8% to 30.3%, suggesting that further scaling remains beneficial. These results demonstrate that small, dedicated models trained from scratch with a language-appropriate tokenizer offer a viable path for low-resource language technology, achieving competitive performance at a fraction of the computational cost. All models and the tokenizer are released under open licenses. --- ### Can ChatGPT Really Understand Modern Chinese Poetry? Authors: Shanshan Wang, Derek F. Wong, Jingming Yao, Lidia S. Chao Published: 2026-03-21 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.20851v1 ChatGPT has demonstrated remarkable capabilities on both poetry generation and translation, yet its ability to truly understand poetry remains unexplored. Previous poetry-related work merely analyzed experimental outcomes without addressing fundamental issues of comprehension. This paper introduces a comprehensive framework for evaluating ChatGPT's understanding of modern poetry. We collaborated with professional poets to evaluate ChatGPT's interpretation of modern Chinese poems by different poets along multiple dimensions. Evaluation results show that ChatGPT's interpretations align with the original poets' intents in over 73% of the cases. However, its understanding in certain dimensions, particularly in capturing poeticity, proved to be less satisfactory. These findings highlight the effectiveness and necessity of our proposed framework. This study not only evaluates ChatGPT's ability to understand modern poetry but also establishes a solid foundation for future research on LLMs and their application to poetry-related tasks. --- ### Glove2Hand: Synthesizing Natural Hand-Object Interaction from Multi-Modal Sensing Gloves Authors: Xinyu Zhang, Ziyi Kou, Chuan Qin, Mia Huang, Ergys Ristani, Ankit Kumar, Lele Chen, Kun He, Abdeslam Boularias, Li Guan Published: 2026-03-21 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2603.20850v2 Understanding hand-object interaction (HOI) is fundamental to computer vision, robotics, and AR/VR. However, conventional hand videos often lack essential physical information such as contact forces and motion signals, and are prone to frequent occlusions. To address the challenges, we present Glove2Hand, a framework that translates multi-modal sensing glove HOI videos into photorealistic bare hands, while faithfully preserving the underlying physical interaction dynamics. We introduce a novel 3D Gaussian hand model that ensures temporal rendering consistency. The rendered hand is seamlessly integrated into the scene using a diffusion-based hand restorer, which effectively handles complex hand-object interactions and non-rigid deformations. Leveraging Glove2Hand, we create HandSense, the first multi-modal HOI dataset featuring glove-to-hand videos with synchronized tactile and IMU signals. We demonstrate that HandSense significantly enhances downstream bare-hand applications, including video-based contact estimation and hand tracking under severe occlusion. --- ### GOLDMARK: Governed Outcome-Linked Diagnostic Model Assessment Reference Kit Authors: Chad Vanderbilt, Gabriele Campanella, Siddharth Singi, Swaraj Nanda, Jie-Fu Chen, Ali Kamali, Amir Momeni Boroujeni, David Kim, Mohamed Yakoub, Jamal Benhamida, Meera Hameed, Neeraj Kumar et al. Published: 2026-03-21 Categories: cs.CV, cs.CE, q-bio.TO arXiv: https://arxiv.org/abs/2603.20848v1 Computational biomarkers (CBs) are histopathology-derived patterns extracted from hematoxylin-eosin (H&E) whole-slide images (WSIs) using artificial intelligence (AI) to predict therapeutic response or prognosis. Recently, slide-level multiple-instance learning (MIL) with pathology foundation models (PFMs) has become the standard baseline for CB development. While these methods have improved predictive performance, computational pathology lacks standardized intermediate data formats, provenance tracking, checkpointing conventions, and reproducible evaluation metrics required for clinical-grade deployment. We introduce GOLDMARK (https://artificialintelligencepathology.org), a standardized benchmarking framework built on a curated TCGA cohort with clinically actionable OncoKB level 1-3 biomarker labels. GOLDMARK releases structured intermediate representations, including tile coordinate maps, per-slide feature embeddings from canonical PFMs, quality-control metadata, predefined patient-level splits, trained slide-level models, and evaluation outputs. Models are trained on TCGA and evaluated on an independent MSKCC cohort with reciprocal testing. Across 33 tumor-biomarker tasks, mean AUROC was 0.689 (TCGA) and 0.630 (MSKCC). Restricting to the eight highest-performing tasks yielded mean AUROCs of 0.831 and 0.801, respectively. These tasks correspond to established morphologic-genomic associations (e.g., LGG IDH1, COAD MSI/BRAF, THCA BRAF/NRAS, BLCA FGFR3, UCEC PTEN) and showed the most stable cross-site performance. Differences between canonical encoders were modest relative to task-specific variability. GOLDMARK establishes a shared experimental substrate for computational pathology, enabling reproducible benchmarking and direct comparison of methods across datasets and models. ---