Robotics
UniWAM Technical Report: Unified Mobile Manipulation via Mixed-Stream World-Action Modeling and Manipulation Anchor Pose Supervision
Wei Xue, Keliang Liu, Mingzhang Cui et al. (+14 more)
Mobile manipulation requires precise
navigation to a manipulation-ready pose followed by reliable object interaction. These two stages differ in action spaces and visual requirements, which complicates unified policy learning. In addition, collecting diverse real-world
navigation data with explicit manipulation-ready pose supervision remains costly and difficult to scale. We introduce UniWAM, a unified mixed-stream world-action model with separate action encoders and output heads for
navigation and manipulation, sharing a common backbone. This design supports joint representation learning on independently sampled
navigation and manipulation data. UniWAM supports independent inference for either stream and batch-parallel inference for both. We further introduce Manipulation Anchor Pose (MAP) supervision for where to stop and how to orient for manipulation. An automated pipeline constructs MAP-Data from large-scale 3D scenes, yielding over 1.5 million episodes and 7,500 hours. MAP-Data provides per-frame target-object bounding boxes and image-plane MAP coordinates as auxiliary
navigation supervision. Together with projected end-effector trajectories for manipulation, these prediction targets provide stream-specific image-plane supervision for action learning from
egocentric observations. With large-scale MAP-Data, UniWAM outperforms the strongest external baselines on our MAP-Bench by 30.1% in position error and 44.0% in heading error. Across 24 real-
robot tasks, UniWAM achieves leading results in MAP
navigation and mobile manipulation, with competitive manipulation performance. We have released code, data, and
benchmark.
Read the full abstract ▾
All authors Wei Xue, Keliang Liu, Mingzhang Cui, Jinhua Xie, Jinjie Wei, Jianan Hou, Jingcheng Lu, Lintao Wang, Kaixiang Qiu, Yizhou Liu, Xinghai Ye, Jinghang Han, Mingcheng Li, Jie Gu, Shunli Wang, Lihua Zhang, Dingkang Yang
Machine Learning
Beyond Simulation: Retain-and-Repair Neural Operators for Real-World Adaptation
Woojin Cho and Junghwan Park
Neural operators increasingly benefit from pretraining on numerical simulations, yet adapting them for real-world prediction remains challenging. We introduce the Retain-and-Repair Neural Operator (R$^2$NO), a framework for adapting simulation-pretrained operators to real-world data while retaining useful pretrained structure. The pretrained operator is first finetuned on real data and then frozen to provide a source prediction, and a shared repair module learns a sequence of refinements from the same observations. Using orthogonal
Fourier projections, a spectral ensemble fits a small ridge regression within each
cell of the
Fourier domain and combines the refinements by weights fitted on a held-out split of the real data. The
cells are defined jointly by radial ranges, angular sectors, and measured channels, allowing refinement depth to vary with frequency magnitude, with orientation, and across channels. Including the source prediction as a candidate makes retention available in every
cell, and independently trained repair modules enter the same combination as additional candidates. On all RealPDEBench systems and six backbones, R$^2$NO consistently outperforms full finetuning and iterative refinement. The framework treats adaptation depth as a
cell-specific choice learned from real data.
Read the full abstract ▾
Machine Learning
When, Not How Much: Evaluating Time-Series Foundation Models on Sparse Events
Daniel Schoess and Florian von Wangenheim
Pretrained time-series foundation models (TSFMs) are evaluated as forecasters of future values, yet for sparse series many decisions depend only on which future periods contain activity. Standard
benchmarks do not assess this. On five sparse
datasets, we rank positions within forecast windows that contain both events and zeros. The released point forecasts of 12 TSFMs improve chance-corrected average precision over training-free references by at most 0.031, and in chance-corrected AUC the median TSFM falls below them on every
dataset. With event supervision, linear probes of six frozen backbones improve on their backbone's point forecast in 29 of 30 backbone–
dataset pairs. Averaging the predicted
quantiles instead of taking their median improves the ranking of most TSFMs that forecast the median, and on two
datasets the strongest such outputs rival the probes. The probes' advantage over raw-context learners depends on the
dataset, and under the same probe, pretrained features outperform randomly initialized ones for five of six backbones. For sparse-event ranking, released point forecasts thus add little over simple references, whereas lightweight event heads on frozen TSFMs rank events better than these forecasts, and the best of them exceed gradient-boosted trees trained on the raw context on three of the five
datasets. More broadly, assessing pretrained forecasters on tasks beyond value
forecasting requires reporting their outputs, supervised probes of their representations, and raw-context and randomized controls side by side, since each supports a different conclusion.
Read the full abstract ▾
Computation and Language
TTLab at Daleel 2026: STAR-Ar, Sequence Tagging for Argument Recognition in Arabic
Bhuvanesh Verma, Ali Abusaleh and Alexander Mehler
Argument Mining (AM) is a critical NLP task that remains significantly under-resourced in Arabic. This paper presents $\testtt{STAR-Ar}$, a BERT-BiLSTM-CRF architecture for argument discourse detection and classification, as our system for Daleel 2026, the inaugural Arabic argument mining shared task. The task requires the identification and classification of argumentative discourse units (ADUs) in debate and editorial texts.We jointly model these two objectives as a
token-level sequence labeling task using a BERT-BiLSTM-CRF architecture that combines contextual
transformer embeddings with structural transition constraints to support accurate span detection. $\testtt{STAR-Ar}$ achieves an F1-score of 72.69 on validation and 73.7 on test data. Our domain-specific analysis shows that models trained exclusively on editorials underperform those trained on debates, a disparity we primarily attribute to the smaller size of the editorial
dataset. The code for $\testtt{STAR-Ar}$ is available at ${\href{https://github.com/ENTAILab/daleel_2026_Arabic-Argumentative-Discourse-Mining}{\faGithub~TTLab at Daleel 2026}}$
Read the full abstract ▾
Robotics
RoboAssist: Interactive Human-Humanoid Planning for Long-Horizon Surgical Assistance
Jingwei Jia, Keyu Zhou, Jiewei Wang et al. (+7 more)
Long-horizon
surgical assistance requires
humanoid robots to coordinate with evolving human activities while maintaining safety across planning and execution. We present RoboAssist, an
agent-based framework for interactive human-
humanoid planning that integrates workflow
reasoning, task coordination, and cross-layer safety. At its core is an asymmetric dual-track representation that separates partially observed human process states from executable
robot task sequences. By updating human-process estimates, scene context, and task dependencies online, RoboAssist revalidates the remaining task sequence and replans only the affected suffix when workflow requests change. A cross-layer safety architecture combines preventive
navigation regulation, reactive regulation during close-range handover, and independent whole-body runtime supervision. This design couples online task coordination with safety constraints throughout execution. We demonstrate the framework on a Unitree G1
humanoid robot in long-horizon, multi-stage simulated
surgical assistance scenarios encompassing
multimodal interaction, instrument handling, medical material transport,
navigation, and safe human-
robot handover. Experiments show multi-stage task completion and adaptation to workflow-request changes. A targeted full-replanning ablation shows that residual replanning reduces plan-update latency and post-update
token usage. Separate safety experiments demonstrate complementary protection across
navigation, handover, and runtime supervision. Additional results and demonstrations are available online at https://roboassist.github.io.
Read the full abstract ▾
All authors Jingwei Jia, Keyu Zhou, Jiewei Wang, Peisen Xu, Xingyuan Zhou, Liang Wang, Jiming Chen, Gaofeng Li, Jin Wang, Shunlei Li
Machine Learning
From Search to Signal: Online Post-Training in Automatic Heuristic Design
Yilun Yuan, Tianyu Zhou and Zhenzhou Tang
Large
language model (LLM)-based automatic heuristic design (AHD) iteratively proposes and refines heuristics, pairing design rationales with executable code. Task-specific evaluators assess programs; execution outcomes and performance scores guide search. Many AHD systems keep the generator frozen; EvoTune and Co-
Evolution of
Algorithms and
Language Model (CALM) instead update it from evaluated candidates. When such outcomes drive reinforcement learning with verifiable rewards (RLVR), they create a search-coupled loop: the evaluated candidate stream supplies both search-state updates and training signals for the model that generates future candidates. Yet validity and performance do not uniquely determine useful model updates; converting them into learning signals must account for the
prompt and evolving search state that produced each candidate. We formulate online post-training of small open-weight LLMs in AHD as context-dependent signal construction and develop alternative mappings from program validity, task performance, and generation context to update signals. Using shared evaluated rollouts and matched update budgets, controlled experiments across AHD tasks and model families compare these mappings with online post-training baselines, testing their effects on validity, performance among valid proposals, and the yield of valid proposals that improve under contextual comparisons. Complementary checkpoint, frozen-search, and live-system evaluations assess whether proposal-level gains appear in updated checkpoint behavior and subsequent search, rather than arising solely from accumulated search state. A resource-matched comparison under pre-specified cost accounting tests whether online updating adds value beyond additional search with a frozen generator. Together, this design avoids treating end-to-end search gains alone as evidence of stronger heuristic-design capabilities.
Read the full abstract ▾
Artificial Intelligence
SkillFM: Generating Skills for LLM Agents via Latent Flow Matching
Zuming Zhang, Jie He, Yizhe Zhang et al. (+1 more)
Textual skills provide reusable guidance for large
language model
agents, but existing approaches often rely on manually curated skill banks or reinforcement learning with indirect and delayed feedback. We introduce SkillFM (Skill Flow Matching), a generative framework that synthesizes task-conditioned textual skills directly without test-time skill
retrieval. Our framework combines a codec for encoding and reconstructing textual skills in a continuous latent space with a conditional flow model trained using improved MeanFlow. At inference time, the learned velocity field enables single-step latent sampling, and an LLM-based decoder converts the sampled representation into textual guidance for a frozen downstream
agent. We evaluate the framework on
embodied tasks, question answering, and web shopping. On ALFWorld and Search-QA, our method achieves the best overall performance among the compared vector-based skill approaches. Our analyses further demonstrate that latent skill generation is an effective alternative to
retrieval-based skill augmentation. Our code and training skill libraries are available at https://github.com/lulushang999/SkillFM.
Read the full abstract ▾
All authors Zuming Zhang, Jie He, Yizhe Zhang, Jeff Z. Pan
High Energy Physics (Phenomenology)
New Observables for Resonant and Non-Resonant New Physics Searches at Colliders
Julien Touchèque
The discovery of the
Higgs boson cemented the Standard Model of particle physics (SM) as we know it and observations have so far confirmed SM predictions. Now, too many beyond-the-Standard-Model (BSM) models can be explored to overcome the remaining shortcomings of the SM. As a result, generic models present a useful intermediary step to point towards the next extension of the SM probing multiple Ultraviolet models simultaneously. In this work, two examples are investigated. First, off-shell effects of potential heavy new particles are provided in the Standard Model Effective Field Theory (SMEFT) through higher dimensional operators. In particular, Charge-parity (CP)
symmetry violating contributions from Wilson coefficients are ranked by using U(1) global flavour
symmetries on CP-odd operators of the Warsaw basis. Asymmetries of simple triple products are proven to be more sensitive than the cross-section to these SMEFT operators and more efficient than most of the usual observables used in experiments. Then, on-shell effects of singlet and octet scalars of SUC(3) are independently searched for in four-top production. The signal consists of the associated production of a scalar with a top
quark pair and of pair production in the case of the octet producing boosted top-
quark pairs. The processes are generated at next-to-leading order in
quantum chromodynamics with a careful renormalisation procedure and the correct selection of loop diagrams. Using improvements in top-jet tagging techniques and in the main background simulation as well as a new technique to select the events, a new particle search is performed to constrain the respective mass and coupling to top
quarks of new resonances.
Read the full abstract ▾
Computer Vision
InfoAgent: Traceable Generation and Repair of Evidence-Grounded Infographics
Yifan Li, Tong Li, Qi Zeng et al. (+6 more)
Reliable infographic generation requires facts, symbols, and visual relations to remain consistent through rendering and revision. Correcting one element also requires tracking its supporting evidence and the dependencies affected by the change. We present InfoAgent, a training-free framework for evidence-bound visual-symbolic program synthesis. Its Infographic Visual Description (IVD) records factual payloads, evidence provenance, execution routes, and verification obligations in a typed dependency graph. Retrieved design priors guide compilation, and layered execution combines raster synthesis with editable symbolic and binding objects while retaining their traces. Dependency-aware repair localizes corrections, rechecks affected dependencies, and requires protected obligations to remain satisfied under the declared checkers. Unresolved obligations remain explicit. On IGenBench, InfoAgent achieves 93.0 Q-ACC and 59.0 I-ACC. We also introduce InfoGraphicBench-Evidence, where complete-checklist pass rates on 200 test requests increase from 21.5% for Same-IVD
Prompt to 23.5% for the initial layered output and 28.5% after repair, using the same evidence and initial IVD. On 120 audited repair cases, localized repair edits 12.4% of the canvas on average, compared with 67.3% for global regeneration.
Read the full abstract ▾
All authors Yifan Li, Tong Li, Qi Zeng, Lishuai Gao, Ruwei Pan, Cong Wei, Shaohua Kevin Zhou, Zhuoliang Kang, Xiaoming Wei
Atmospheric and Oceanic Physics
Butterfly Effect Confirmed in Global AI Weather Models: Evidence from Tropical Cyclone Forecasting
Jeremy Cheuk-Hin Leung, Daosheng Xu, Weiye Yu et al. (+9 more)
A paradox recently emerged in artificial intelligence (AI) weather prediction research. While some claim AI weather models cannot simulate atmospheric butterfly effect, this conflicts with AI models' limited predictability and advances in AI ensemble
forecasting. This study demonstrates via counterexamples that the butterfly effect does exist in AI weather predictions. For Super Typhoon Khanun, AI predictions are constrained by a double-attractor system. Minor initial perturbations confined to two regions trigger state transitions between two local attractors, causing a 1006-km difference in the predicted storm position on Day 7. This behavior is consistent with numerical weather prediction models and observed in ~12% of tropical cyclones in the past 5 years. These findings verify AI's ability to capture atmospheric chaos and provide the physical basis for AI ensemble
forecasting.
Read the full abstract ▾
All authors Jeremy Cheuk-Hin Leung, Daosheng Xu, Weiye Yu, Shaojing Zhang, Xiaodong Zeng, Gaozhen Nie, Jie Feng, Jingchen Pu, Yi Li, Kaijun Ren, Qingcun Zeng, Banglin Zhang
Computer Vision
Shulin Tian, Junsu Kim, Shuai Liu et al. (+17 more)
Real-world
embodied tasks, from everyday activities to professional procedures, require
agents to act under physical constraints while tracking evolving object and task states. Tool use sits at the heart of such tasks, as many everyday and professional activities are tool-mediated. Understanding them requires
reasoning about affordances, hand-tool-object geometry, procedural progress, and causal effects on target objects. Yet despite strong performance on perception-oriented
video tasks such as captioning and general
video QA, current
multimodal video models remain limited in this form of tool-centric
embodied reasoning. Progress in this direction has been limited by the lack of real-world
egocentric data and diagnostic
benchmarks. To address this gap, we introduce EgoTools, the first comprehensive suite for
egocentric tool-use understanding. It consists of two complementary components: EgoTools-Data, a large-scale corpus of 100 hours of tool-centric
egocentric recordings with synchronized audio, dense captions,
reasoning-heavy narrations, and supplementary 3D information; and EgoTools-Bench, a diagnostic
benchmark of 1,000 QA pairs across four tracks that cover tool-use understanding from perception and geometry to procedure and causal
reasoning. Experimental results show that current models still struggle to ground tool use in visual evidence: Gemini-3.1-Pro achieves 66.9% overall accuracy but only 51.7% on Perception & Grounding. Beyond evaluation, we validate EgoTools-Data as a training resource. On the full 1,000-question
benchmark, full supervised fine-tuning improves Qwen3-VL-8B-Instruct from 50.0% to 60.9%, under strict source-
video separation. Together, these results establish EgoTools as a unified resource for both training and diagnostic evaluation of real-world
egocentric tool-use understanding.
Read the full abstract ▾
All authors Shulin Tian, Junsu Kim, Shuai Liu, Hao Li, Yujiao Shen, Sihan Li, Zhe Yang, Yeongon Kim, Feiyu Li, Jialin Wu, Yichi Zhang, Wenhui Wang, Runmao Yao, Yuhao Dong, Zhaoxi Chen, Fangzhou Hong, Antonino Furnari, Jingkang Yang, Hongyuan Zhu, Ziwei Liu
Adaptation and Self-Organizing Systems
Recovery Random Walks for Extreme Events on Complex Networks
Karan Singh, Narendran R. V., V. K. Chandrasekar et al. (+1 more)
Extreme events are widely studied within simple random walk frameworks, where their probability is determined by the network structure and stationary walker distribution. Here, we propose a recovery random walk (RRW) model in which extreme events temporally `freeze' the nodes where they occur for a fixed duration $\Delta$ ($\Delta=0$ recovers the original model), trapping walkers and reducing the effective mobile population, thereby making the model more practical. We derive a first-principles description of this feedback and a delayed differential equation for the frozen-node fraction. This finite freezing produces an initial overshoot, followed by damped oscillatory relaxation to a steady state for the fraction of the frozen nodes. We find that freezing suppresses extreme event probability while preserving its degree dependence dynamics, and suppresses the EE frequency. Further, the analytical prediction for the fraction of the frozen nodes agrees closely with simulations. This model closely reflects real-world scenarios, yielding more practical EE statistics.
Read the full abstract ▾
All authors Karan Singh, Narendran R. V., V. K. Chandrasekar, D. V. Senthilkumar
Plasma Physics
Phase-Space Non-Integrability of Alpha Particles in Near-Omnigeneous Stellarators
A. L. Lachmann, A. R. Knyazev and E. J. Paul
In a burning
plasma, fusion-born alpha particles sustain the fusion reaction by heating the
plasma core. Alpha-particle confinement in stellarators is sensitive to both deviations from omnigeneity and time-dependent Alfvénic perturbations. Traditional Poincaré maps rely on
symmetry assumptions or dimensionality reductions that fail for stellarators with quasisymmetry (QS) error or multi-harmonic Alfvénic perturbations. To address these challenges, we investigate weighted Birkhoff averaging (WBA) as a diagnostic to detect the onset of chaos in guiding-centre trajectories subject to general
magnetic perturbations. WBA exploits the super-convergence of time averages along quasiperiodic orbits, discriminating regular from chaotic trajectories using any smooth observable rather than an exact invariant of motion. Applying WBA to the canonical momentum $P_\eta$, we show that convergence tracks underlying orbit regularity even when quasisymmetry is imperfect and $P_\eta$ is not strictly conserved. Convergence is measured by the digit accuracy of the weighted average along guiding-centre orbits in fields with imperfect omnigeneity, with and without Alfvén eigenmodes computed using AE3D. We use this diagnostic to map phase-space chaos as a function of
equilibrium invariants to identify transport barriers in unperturbed and perturbed stellarator
equilibria. This work demonstrates that chaos induced by many Alfvén harmonics is often not apparent from the largest harmonic alone. We also find evidence that the onset of chaotic behaviour is sensitive to the local radial profile of the wave and not only to its amplitude at the resonance.
Read the full abstract ▾
Robotics
Beyond the Current Scene: Event-Referential Grasping with Active View Selection
Hyunjoon Lee, Haebeom Jung, Eunsung Cha et al. (+4 more)
A
robot that observes people interacting with objects should be able to carry out later requests that refer back to those interactions. Such requests may specify a grasp target by the role it played in a past event rather than by its name or appearance. Moreover, the target may no longer be visible when the
robot is asked to act. We present BeyondSCe, a zero-shot robotic grasping system for this event-referential setting. Given the event history and the current scene, the system identifies the requested object or part and localizes it for grasping. If the target is occluded, it combines an event prior recovered from the history with current scene geometry to select camera viewpoints likely to reveal the target. The system uses pretrained models without additional task-specific training. In real-
robot experiments with a single wrist-mounted RGB-D camera, it achieves grasp success rates of 76% and 77% for initially visible and occluded targets, respectively, compared with 40% and 55% for the strongest baseline in each condition. On four additional scenes with heavy occlusion, it increases grasp success rates from 75% to 95% while reducing the mean number of views from 3.35 to 2.20, compared with an active-perception baseline given the target's ground-truth 3D bounding box.
Read the full abstract ▾
All authors Hyunjoon Lee, Haebeom Jung, Eunsung Cha, Daeun Lee, Yu-Chiang Frank Wang, Jaesung Choe, Jaesik Park
Machine Learning
Wavelet Flow Matching for Time Series
Lucas Poinsignon, Jorge da Silva Goncalves, Samuel Ruiperez-Campillo et al. (+1 more)
Synthetic time series are increasingly used for data augmentation,
privacy-preserving data sharing, and downstream model development, yet faithfully reproducing both multi-scale temporal structure and cross-channel dependencies remains challenging. We study multivariate time-series generation through flow matching in the wavelet domain. By operating on multilevel discrete wavelet coefficients rather than directly in the time domain, the model represents coarse structure and progressively finer details at separate scales. Their naturally different variances further induce an implicit coarse-to-fine generative process without requiring an explicit multi-scale schedule. Since the transform acts independently on each channel, we pair it with a channel-
token transformer whose
attention directly models cross-channel dependencies. Across seven
benchmark datasets and four sequence lengths, our method is best or tied on a majority of
dataset-metric combinations, with the largest and most consistent improvements in Context-FID and discriminative score.
Read the full abstract ▾
All authors Lucas Poinsignon, Jorge da Silva Goncalves, Samuel Ruiperez-Campillo, Julia E. Vogt
Information Theory
Dynamic Coded Distributed Computing with Joint File Placement and Reduce Function Reassignment
Chenglin Li, Minquan Cheng and Youlong Wu
Coded distributed computing (CDC) mitigates the Shuffle communication bottleneck through coded multicast, but existing schemes with predetermined worker populations do not directly adapt to worker arrivals and disconnections. To address these dynamics, we propose a dynamic CDC scheme that preserves surviving workers' caches and Reduce assignments while exploiting the storage brought by arriving clusters. The model accommodates general MapReduce tasks under a placement that is partly fixed and partly new. The scheme uses Cartesian products of cluster-wise placement delivery arrays (PDAs) for file placement. A delivery-aware greedy
algorithm reassigns abandoned Reduce functions, and a Communication Array allows workers with no Reduce function to serve as coded transmitters. For arbitrary feasible disconnections, the scheme's exact communication load is derived. A converse bound under one-shot communication shows that, in the absence of disconnections, the scheme lies within a factor of 2 of the lower bound under these assumptions when new clusters arrive, and within a factor of 4 otherwise.
Read the full abstract ▾
Quantum Physics
Efficient Learning of Fermionic Magic States under Free-Fermion Evolution
Jiwon Heo, Myeongjin Shin and Changhun Oh
We establish efficient learning for a family of fermionic magic states under unknown number-conserving free-
fermion evolution. Each input block has a definite particle number and is a superposition of Fock states, with occupied mode sets disjoint both within and across blocks. The key idea is to exploit the spectral structure of particle reduced density
matrices (RDMs) to separate contributions from individual blocks from those involving several blocks, allowing us to reconstruct the hidden block structure. For a fixed upper bound on the particle number per block, our
algorithm uses single-copy measurements and
polynomial sample and classical computational complexity to recover a compact description of the state with prescribed fidelity and high probability, without prior knowledge of the block decomposition or the
evolution. RDMs up to this upper bound suffice for reconstruction. We further show that this RDM order is necessary in general: two orthogonal states in the family can have identical RDMs at every lower order. These results show that an extensive number of non-Gaussian blocks can be compatible with efficient state learning.
Read the full abstract ▾
Artificial Intelligence
EHR-RobustGym: Benchmarking and Training Agents for Robust Clinical Reasoning
Yitong Qiao, Yancheng Jin, Lei Liu et al. (+4 more)
In hospital workflows, electronic health records (EHRs) are often noisy, and may not contain the evidence needed to confirm events or measurements referenced in a clinical query. Even when database
retrieval succeeds, clinical
agents can overlook such discrepancies and return plausible but unsupported answers. We introduce EHR-RobustGym, a scalable and interactive environment for evaluating and training robust clinical
agents grounded in noisy EHRs. Built on MIMIC-IV hospital records (365K patients, 31 tables, and over 500M records), EHR-RobustGym comprises 5,486 Clean-Noise pairs spanning six clinical intents and both patient-level and population-level queries. The pairs test robustness to Record-level, Value-level, and Query-level noise, while interactive SQL/Python execution and outcome verification support trajectory collection and training. Evaluating multiple LLMs reveals substantial robustness gaps: average task success across proprietary and large-scale open-weight models drops from 62.2% on Clean questions to 37.9% on Noise questions. At k=4, pass^k consistency falls below 50% for most evaluated models, exposing instability in clinical task completion. Supervised fine-tuning and reinforcement learning in EHR-RobustGym improve performance, with gains generalizing to five external EHR
benchmarks. Together, these results position EHR-RobustGym as a testbed for evaluating and improving the evidence-grounded robustness of clinical
agents.
Read the full abstract ▾
All authors Yitong Qiao, Yancheng Jin, Lei Liu, Yue Shen, Jian Wang, Jinjie Gu, Zhixuan Chu
Group Theory
The Margolis-Rhodes Monoid of a Graph
Stuart Margolis and John Rhodes
We investigate the structural, combinatorial, ideal-theoretic and Krohn-Rhodes complexity of the Margolis-Rhodes monoid, MR(G), of a finite simple graph G, viewed topologically as a 1-dimensional simplicial complex. Alongside the full monoid, we examine some subsemigroups including St(G), defined by the condition that the full inverse image is an edge or the empty set and Inj(G), the monoid of all partial 1-1 continuous functions. We provide explicit combinatorial enumerations and struture for paths and cycles. We compute Green's relations showing in particular that the partial order of regular J-classes is isomorphic to the poset of induced subgraphs of G. Finally, we apply these structural invariants to Krohn-Rhodes complexity theory. It is known that the Margolis-Rhodes monoid has complexity at most 2 and 1 for St(G) and Inj(G). We show that for cycles the complexity of its Margolis-Rhodes monoid is 2 if and only if the cycle is of length at least 4. For paths, we prove that the complexity of its Margolis-Rhodes monoid is 2 if the path length is at least 13.
Read the full abstract ▾
Computation and Language
Exploring Heterogeneous Model Merging Approach for Complex Knowledge Transfer
Jiahe Fan, Si Chen, Yinghao Hou et al. (+4 more)
Specialized models encode task-oriented behavior, but transferring that behavior to a general
language model usually requires training,
distillation, or representation
alignment. We study whether such ability can instead be transferred directly at the parameter level. We apply two existing training-free heterogeneous merging methods, previously shown to transfer knowledge between general
language models, to specialist-to-general transfer, projecting a specialist donor into the recipient's shape and interpolating backbone parameters without gradient updates or semantic
alignment. Intersection-Merge (IM) injects a prefix-aligned donor slice matching the recipient shape, while Activate-Prune-Merge (APM) uses forward-pass activation statistics to select which donor dimensions to retain before injection. Across
embedding, reranking, reward modeling, and MoE code-specialist transfer, both methods improve the general recipient, showing that simple heterogeneous merging can move capabilities across diverse specialist roles.
Read the full abstract ▾
All authors Jiahe Fan, Si Chen, Yinghao Hou, Wenbo Xia, Ke Xu, Hong Xie, Enhong Chen
Computation and Language
Making Grid Beam Search Less Greedy
Sean Papay and Roman Klinger
A common formalism for constraining the output of autoregressive text generation models involves lexical constraints, words or phrases which are required to occur in the generated text. DFA-constrained beam search and grid beam search are two widely used paradigms for decoding from autoregressive models while enforcing lexical constraints. As the former approach requires a number of forward passes exponential in the number of constraint
tokens, it is often dispreferred to the latter, which requires only linearly many forward calls. However, while grid beam search achieves an exponential speedup, it does so in a manner which does not treat all of the constraints equally. In this paper, we demonstrate that grid beam search is biased to incorporate easier-to-satisfy constraints first, leaving harder constraints to the end of the sequence. This contrasts with DFA-constrained beam search, which exhibits no such bias. To address this shortcoming, we propose fair grid beam search, a modification to grid beam search which avoids this bias while still requiring only linearly many forward passes. Experimentally, we confirm grid beam search's bias on two constrained generation tasks, finding significant differences in how it orders constraint
tokens as compared to DFA-constrained beam search and fair grid beam search. Furthermore, we find that fair grid beam search not only fixes grid beam search's bias, but finds higher-probability strings in the process.
Read the full abstract ▾
Machine Learning (Statistics)
A Dynamical Theory of LoRA in Continual Learning
Théo Marchetta, Filippo Alessandroni, Alessandro Breccia et al. (+2 more)
Despite the widespread use of Low-Rank Adaptation (LoRA), little is known about its dynamics in continual learning and the mechanisms by which low-rank updates affect catastrophic forgetting. We provide an asymptotically exact dynamical characterization of LoRA in a solvable two-task teacher-student model. In the high-dimensional online-learning limit, we derive a closed system of ordinary differential equations for a finite set of macroscopic order parameters, yielding exact expressions for the generalization errors throughout both the initial Task 1 learning phase and the subsequent LoRA fine-tuning on Task 2. The theory quantitatively matches finite-dimensional simulations and exposes two characteristic effects of LoRA: low-rank adaptation reduces interference with features learned on the first task, but its initialization slows adaptation to the second task. Building on this mechanistic picture, we analyze a state-dependent masking strategy that freezes hidden units carrying the strongest first-task representations and restricts adaptation to the complementary subspace. This structural partitioning markedly reduces forgetting, while preserving plasticity on the new task. Our framework further clarifies the role of adapter rank: transfer improves only up to the intrinsic dimensionality of the target task and saturates beyond it, while forgetting continues to grow with rank. These results provide a dynamical and
geometric account of how low-rank adaptation organizes information across sequential tasks and are qualitatively reproduced on a sequential MNIST
benchmark.
Read the full abstract ▾
All authors Théo Marchetta, Filippo Alessandroni, Alessandro Breccia, Alessandro Ingrosso, Federica Gerace
Computer Vision
COBICount: Separating Object and Background Responses for Remote Sensing Object Counting Without Training on Target Data
Junjing Zheng, Zhiyi Zhou, Ningrui Yang et al. (+1 more)
Remote sensing object counting estimates how many buildings, vehicles, or ships appear in overhead images. Most supervised counters predict a density map, whose sum gives the object count, and assume similar categories, sizes, and backgrounds. Applying them across regions, sensors, or categories often requires target data or further training, which may be costly or unavailable. We study source-only counting. Training for the counting task and model selection use one group of images that shares an object category and similar imaging conditions, with one point marking each object. Target images and information remain unavailable until the model is fixed. This reduces data preparation but makes transfer harder. A model trained on one source may place high density values, called responses, on real objects and repeated background structures. Road edges, parking grids, roof boundaries, and water boundaries may then be counted as objects, creating candidate origin ambiguity. COBICount separates response generation, acceptance, and background suppression. Candidate Evidence (CE) generates possible responses. Candidate Acceptance (CA) keeps compact responses centered on objects. Bias Isolation (BI) reduces responses associated with repeated background structures. Their outputs form the final density map. Trained on RSOC Building and evaluated directly on DOTA Large Vehicle, Small Vehicle, and Ship, COBICount achieves the lowest mean absolute error (MAE) averaged over the target domains among the compared methods, 174.132. It uses 5.07 million parameters and 17.41 billion floating point operations for a 512x512 input. COBICount improves transfer without target data or training for each target. The code will be available at: https://github.com/yixuxi22/COBICount.
Read the full abstract ▾
All authors Junjing Zheng, Zhiyi Zhou, Ningrui Yang, Hongying Meng
Computation and Language
Ready2Blend: From Natural-Language Instructions to Composable Alignment Prompts
Jeesu Jung, Hwan Chang, Juseon Do et al. (+5 more)
Continual
alignment requires LLMs to adapt to new requirements without forgetting previously acquired behaviors. Natural-
language instructions are flexible and composable but offer only indirect control, whereas post-training provides stronger adaptation at the cost of repeated parameter updates. We introduce Ready2Blend, which combines the flexibility of natural
language with learned
alignment. AlignFormer maps each requirement to a fixed-length
alignment prompt stored in a modular
prompt bank, while the backbone and prior
prompts remain frozen. Composability regularization transfers the semantic geometry of textual requirements into
prompt space, enabling inference-time blending and reweighting. Across two practical continual
alignment settings, Ready2Blend is the only frozen-backbone method that matches post-training-based
alignment methods, reaching $93.1$-$98.5\%$ of a joint-training reference with competitive retention, while requiring only a few
prompt tokens and up to $4.3\times$ less training time. Its modular design further enables weighted personalization and order-free composition without retraining. Code will be released upon acceptance.
Read the full abstract ▾
All authors Jeesu Jung, Hwan Chang, Juseon Do, Jeonghwan Choi, Jinho Choo, Sungwoo Nam, S. K. Hong, Hwanjun Song
Fluid Dynamics
Imperfect bifurcations in a laminar 3D bluff body wake: effect of pitch and yaw
Atharva Lagwankar, Alessandro Chiarini, François Gallaire et al. (+1 more)
We study the effect of pitch and yaw on the bifurcations of the laminar flow past an Ahmed body, a bluff body of width-to-height ratio $W/H=1.2$ and length-to-height ratio $L/H=3$. When perfectly aligned with incoming flow, its wake first undergoes pitchfork bifurcations leading to a static vertical or horizontal deflection, and becomes oscillatory via a secondary
Hopf bifurcation at larger Reynolds number $Re$. As commonly observed for imperfect pitchfork bifurcations, any small misalignment preselects one of the steady states, referred to as the "primary" branch, at low $Re$. At larger $Re$, a disconnected "secondary" branch is born via a saddle-node bifurcation. Here, we investigate the effect of misalignment using three-dimensional computations. We first focus on pure pitch and pure yaw with linear stability analysis. In both cases, the secondary branch becomes unstable as soon as the incidence exceeds a small value. In the pure yaw case, the primary branch becomes unstable via a horizontal
Hopf bifurcation, destabilised by yaw. In the pure pitch case, the primary branch sees the crossover of vertical
Hopf and horizontal pitchfork bifurcations, destabilised and stabilised by pitch, respectively. This motivates use of DNS to study this nonlinear competition, already observed for aligned Ahmed bodies in ground proximity. We then study the effect of simultaneous yaw and pitch with a weakly nonlinear analysis. The competition between the two stationary modes gives rise to rich bifurcation diagrams, allowing us to predict the number of stable solutions and phase space trajectories between different states
Read the full abstract ▾
All authors Atharva Lagwankar, Alessandro Chiarini, François Gallaire, Edouard Boujo
Computer Vision
Rethinking Multi-Image Re-Representation in Multi-Image Understanding
Gengyuan Zhang, Xiao Han, Xinyu Xie et al. (+2 more)
Multi-image understanding requires MLLMs not only to recognise the content of individual images, but also to organise visual evidence distributed across them. We study this problem through multi-image re-representation, viewing prompted Chain-of-Thought
reasoning and agentic visual tool use as different ways of re-organising visual evidence during
reasoning. We introduce Mosaic, a general-purpose multi-image visual harness that enables an MLLM to actively construct visual intermediates with ten composable image operations. We compare five re-representation settings on existing multi-image
benchmarks and on MosaicBench, a new grounding-focused
benchmark for fine-grained multi-image understanding. Our experiments show that the relative benefits of textual and visual re-representation are strongly task-dependent. Visual re-representation is particularly effective for tasks requiring precise visual evidence, including hypothesis testing, precision comparison, and orientation-sensitive
reasoning, while tasks dominated by higher-level semantic content show smaller or less consistent gains. Building on this finding, we train MosaicAgent-8B to use Mosaic with reinforcement learning using only accuracy and format rewards. Without demonstration trajectories or rewards for specific tool-use, the
agent learns to compose visual operations over multiple steps and exhibits diverse problem-solving patterns unpromptedly. Code and data will be released at https://github.com/gengyuanmax/Mosaic.
Read the full abstract ▾
All authors Gengyuan Zhang, Xiao Han, Xinyu Xie, Tong Liu, Volker Tresp
Cryptography and Security
Link Inference Attack on Privacy-Preserving Knowledge Graphs
Emna Bouguerra, Ibtissam Harrouche, Ferran Alborch et al. (+1 more)
Knowledge Graphs (KGs) are widely used to store and share structured information across sensitive domains such as healthcare, fi- nance, and social networks. A common
privacy practice is to delete sen- sitive relations before publishing the graph, under the assumption that removing edges is sufficient to prevent their recovery. In this paper, we challenge this assumption and show that even when a relation is fully or partially hidden, its existence leaves structural traces in the public graph that can be exploited to recover it with high accuracy. To this end, we propose a link inference attack that operates on the
topology of the public graph, and evaluate it under two
privacy scenarios that differ in how the adversary exploits the knowledge available to him. In the first setting where the adversary exploits all topological information, the attack achieves near-perfect discrimination (AP = 0.949, ROC-AUC = 0.999), while in the more realistic one where the adversary makes use of some semantic information, it recovers up to 74% of hidden edges. Build- ing on these results, we further conduct a structural analysis to identify which topological properties of the graph drive the attack success, re- vealing that
privacy risk is not uniform across entities and that certain structural patterns make specific relations significantly more vulnerable to inference than others.
Read the full abstract ▾
All authors Emna Bouguerra, Ibtissam Harrouche, Ferran Alborch, Melek Önen
Machine Learning
LampAttention: Look-Ahead Mixed-Precision FlashAttention for Dedicated Accelerators
Stanislav Budzinskiy, Marian Gloser, Tolunay Yilmaz et al. (+5 more)
While most
attention logits can be computed in low precision without degrading numerical stability, current
attention kernels fail to exploit this phenomenon. We introduce a novel hardware-
algorithm co-design in the form of mixed-precision FlashAttention. Our method accumulates key-query products and evaluates their exponentials in 8-bit formats, then adaptively identifies sensitive sub-blocks and recomputes them in 16-bit formats. We propose the specifications for a dedicated accelerator capable of executing this pipeline efficiently. Simulated experiments with Qwen3 and Gemma 3 show that rerouting a selective minority of sub-blocks to high precision is sufficient to recover the baseline model performance.
Read the full abstract ▾
All authors Stanislav Budzinskiy, Marian Gloser, Tolunay Yilmaz, Ying Hong Tham, Yuanyi Lin, Wenyi Fang, Fan Wu, Philipp Petersen
Artificial Intelligence
Autoresearch in Mixed-Integer Linear and Nonlinear Programming
Yuwei Gu, Yaoxin Wu, Tong Guo et al. (+2 more)
Despite recent progress in autoresearch, applying it to practical operations research problems, typically formulated as NP-hard mixed-integer linear or nonlinear programs (MILPs or MINLPs), remains challenging because effective research requires systematically managing competing ideas and long-horizon experimental trajectories. We introduce AutoMIP, a reusable
agent skill for organizing long-horizon autoresearch in mixed-integer programming through idea pooling and
algorithm tree search. AutoMIP maintains a persistent pool of complementary candidate ideas while organizing executable experiments into an
algorithm tree, enabling the
agent to preserve unexplored hypotheses, refine promising
algorithms, and switch to alternative methodological directions based on historical states. On MILP and MINLP
benchmark cohorts, AutoMIP achieves the highest final success rates among the evaluated autoresearch frameworks. On MIPLib, AutoMIP discovers new best solutions for 31 of 60 instances, surpassing existing autoresearch frameworks. On MINLPLib, it achieves new best solutions for 52 of 60 instances. Ablation studies further demonstrate the complementary contributions of idea pooling and
algorithm tree search, highlighting the importance of jointly maintaining diverse research ideas and structured experimental trajectories for long-horizon autoresearch.
Read the full abstract ▾
All authors Yuwei Gu, Yaoxin Wu, Tong Guo, Wen Song, Zhiguang Cao
Distributed Computing
Communication-Efficient $(1+\varepsilon)\Delta$-Edge Coloring and Lovász Local Lemma
Yi-Jun Chang, Nima Dolatabadi and Hung Thuan Nguyen
We study edge coloring in the two-party edge-partition model, where Alice and Bob each know part of the edge set and must jointly produce a proper coloring with little communication. Previous work gave a deterministic $(2\Delta-1)$-edge-coloring protocol using $O(n)$ bits, leaving open whether fewer colors can be obtained efficiently. We simultaneously reduce both the number of colors and the communication. For every fixed $\varepsilon>0$ and all sufficiently large $\Delta$, we give a public-coin Las Vegas protocol that finds a proper $(1+\varepsilon)\Delta$-edge coloring using $O(ne^{-\gamma\Delta} + 1)$ expected bits and $O\left(\frac{\log n}{\Delta}+1\right)$ expected rounds, where $\gamma>0$ depends only on $\varepsilon$. Thus, the expected communication is $o(n)$ when $\Delta=\omega(1)$ and $O(1)$ when $\Delta\ge C_\varepsilon\log n$, for a sufficiently large constant $C_\varepsilon$. Using only private coins adds $O(\log n)$ expected bits. Our key idea is a new randomized coloring procedure that allows Alice and Bob to color their edges using essentially the same color space with only a small amount of coordination, so most of their random choices remain private. To make this procedure succeed, we develop a communication-efficient constructive Lovász local lemma (LLL) for two parties. Our two-party constructive LLL is also of independent interest. We illustrate its broader applicability by applying it to standard LLL formulations of several other classical problems, obtaining communication-efficient two-party protocols.
Read the full abstract ▾