Peer Revenue Impact factor: volatile

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Tuesday 13 October

Machine Learning

Confine to the Task, Anchor the Rest: Concept-Guided Retention Regularizer for Parameter-Efficient Fine-Tuning

Ibna Kowsar, Ivan Medri, Fanyu Meng et al. (+1 more)

Parameter-efficient fine-tuning (PEFT) has been widely used, yet the adapted model suffers catastrophic forgetting of pre-trained knowledge. While most existing work hardens the adapter by constraining where the weight update may lie, we note that constraining this weight channel does not constrain the activation channel. In this paper, we propose Concept-Guided Retention (CGR), a plug-and-play regularizer that constrains both for any PEFT base. CGR reads a frozen sparse autoencoder (SAE) and treats two disjoint concept sets differently. It softly confines the update to the task-concept subspace (via weight channel), and anchors a selected set of general-purpose concepts to the frozen model (via activation channel). We also provide anchor value bounds based on weight updates to understand how they interact, and show that both terms are necessary. On a CLIP vision panel, CGR significantly improves retention at comparable accuracy for every base, and the same pattern holds for world-knowledge retention under Llama-3.1-8B instruction tuning. Because the protected directions are named SAE concepts, CGR remains inspectable and controllable.
Read the full abstract

All authors Ibna Kowsar, Ivan Medri, Fanyu Meng, Xin Liu

Data Structures and Algorithms

Fully Dynamic Random-Walk Quantities on Undirected Expanders and $\kappa$-Uniform Directed Graphs

Richard Peng, Sushant Sachdeva, Junzhao Yang et al. (+1 more)

We study fully dynamic data structures for random-walk quantities on undirected expanders and fast-mixing directed graphs. For an unweighted graph whose normalized Laplacian has spectral gap at least $\gamma$, we give a $(1\pm\varepsilon)$ effective-resistance data structure with expected update time $O(\gamma^{-4}\varepsilon^{-2}\log n\log^4(n/\varepsilon))$ and expected $O(1)$ query time. For directed graphs, somewhat surprisingly, the OMv conjecture precludes a fast data structure based on rapid mixing conditions alone. To address this, we introduce the notion of $\kappa$-uniform directed graphs, which requires the stationary distribution to $\kappa$-approximate the uniform distribution, and show that in this setting, random walk based data structures with $\text{poly}(\kappa, \log{n}, \varepsilon^{-1})$ update/query times are possible.
Read the full abstract

All authors Richard Peng, Sushant Sachdeva, Junzhao Yang, Yibin Zhao

Cryptography and Security

Local Leakage Resilience of Shamir Secret Sharing: Below Half Threshold Rate

Hai H. Nguyen

We study the local leakage resilience of Shamir secret sharing, where an adversary leaks a bounded number of bits from every share. For $n$ parties and reconstruction threshold $k$ over a prime field $\mathbb{F}_p$ with $p>n$, we prove the following. (1) There is an absolute constant $c_{\mathrm{bh}}<1/2$ such that, for $k\ge c_{\mathrm{bh}}n$, Shamir sharing is resilient to arbitrary one-bit leakage from every share. The best previous threshold was $0.668n$. (2) For every fixed $0<\delta<1/2$ and $k\ge(1/2+\delta)n$, Shamir sharing is resilient to $\left\lfloor\frac{\delta}{24}\log_2 n\right\rfloor$ bits of arbitrary local leakage per share. By a recent attack, this number of bits is optimal up to a constant factor depending on $\delta$. (3) For every fixed threshold rate $c>0$ and $k\ge cn$, Shamir sharing is resilient to quadratic-residue leakage and to almost all local leakage functions of up to $\eta_c\log_2 p$ bits, for a constant $\eta_c>0$. All three results follow from one analytic framework, which we develop for generalized Reed–Solomon codes: local leakage from a uniform codeword is statistically close to the same leakage from independent uniform field elements. The framework expresses the leakage distance as the norm of a function on the leakage outputs. The Efron–Stein decomposition splits this function into orthogonal components indexed by supports of dual codewords. Two tools bound the components: list recovery of the dual code, which is ordinary list decoding for one-bit leakage, and a generalized von Neumann inequality for MDS systems of linear forms that retains linearly many Gowers norms. The first tool gives the result above half, the second gives the constant-rate result, and applying both to different ranges of supports in the same decomposition gives the result below half.
Read the full abstract

Artificial Intelligence

Adaptive-Group Process Rewards for GUI Agents

Zirui Shang, Yang Liu, Xin Shu et al. (+3 more)

Training graphical user interface (GUI) agents with reinforcement learning suffers from sparse outcome rewards, leading to inefficient learning. Process rewards offer a promising direction for credit assignment to intermediate actions; however, accurately evaluating an action requires identifying evidence across long-horizon GUI trajectories, as its value may depend on information beyond its immediate state and observation. In this paper, we propose an adaptive-group process reward framework that dynamically partitions a GUI trajectory into semantically coherent subgoal groups, enabling process reward estimation at both the action and group levels. At the action level, a critic evaluates each action based on the visual changes and subsequent actions within the group for action correctness and contribution to the subgoal. At the group level, the framework aggregates action-level evaluations to assess subgoal completion and contribution to the overall task, and further refines these assessments by leveraging the context of all groups in the trajectory. The process reward of each action is obtained by combining its action-level evaluation with the refined group-level evaluation. We further augment GRPO's outcome advantage with the proposed process rewards to provide finer-grained guidance for GUI agent training. Experiments on process reward benchmarks and downstream RL training demonstrate the effectiveness of the proposed framework.
Read the full abstract

All authors Zirui Shang, Yang Liu, Xin Shu, Zhi Gao, Xinxiao Wu, Lifeng Fan

Machine Learning

Where Small Mistakes Matter: Quantifying State Criticality for Imitation Learning

Yusuf Umut Ciftci, Joonwon Kang, Ajay Suresha Sathya et al. (+1 more)

Imitation learning typically treats every demonstration state as equally important, optimizing a behavior-cloning objective uniformly across the dataset. Yet not all states are created equal: in a small number of critical states, even a slight deviation from the expert action can lead to a failure, while in many other states such errors are recoverable and inconsequential. We formalize state criticality in imitation learning as the worst-case degradation in downstream value induced by a bounded perturbation of the expert action at a given state. We then propose SCRIM, a practical State CRIticality Measure that estimates this quantity by training an adversarial disturbance policy and a value critic around a fixed imitation policy. SCRIM assigns each demonstration state a score reflecting how vulnerable the task outcome is to small, policy-relevant action errors. We show that these scores can be used to reweight behavior-cloning losses, improving imitation at consequential states and increasing overall task performance. We further show that SCRIM enables data curation by selecting compact, high-value subsets of demonstrations that preserve performance while substantially reducing the amount of training data. We evaluate SCRIM across various simulated and real-world robotic manipulation tasks, and demonstrate that SCRIM improves policy training, enables efficient data selection, and provides interpretable insights into which moments in a demonstration are most critical for task success. Project website: https://sia-lab-git.github.io/SCRIM/
Read the full abstract

All authors Yusuf Umut Ciftci, Joonwon Kang, Ajay Suresha Sathya, Somil Bansal

Quantum Physics

QEDIA: Quantum Program Verification in Lean

Yuanjie Ren, Zhicheng Zhang, Junyi Liu et al. (+3 more)

Practical quantum programs interleave quantum and classical computation: classical data shape the quantum operations that are performed, and measurement outcomes shape the classical computation that follows. Formal verification of such programs in a proof assistant requires mechanizing a program logic that handles this interplay consistently at every level, from the mathematics of quantum mechanics, through program syntax and semantics, to specifications and proof rules. Such mechanizations, along with the program proofs built on them, typically require substantial manual effort. We present QEDIA, a framework for verifying quantum programs with classical variables in Lean 4. At its foundation is an open library that formalizes the quantum Hoare logic QHL+ across all these levels and proves the logic sound for both partial and total correctness. We construct the library and the program proofs, under human supervision, using an LLM-based prover agent, whose general-purpose core we extend for quantum program verification. We illustrate QEDIA with verified programs including the quantum Fourier transform, phase estimation, repeat-until-success circuits, and Hamiltonian simulation.
Read the full abstract

All authors Yuanjie Ren, Zhicheng Zhang, Junyi Liu, Jinzheng Li, Yidi Qi, Mingsheng Ying

Machine Learning

Focus on What You Need: Global-to-Local Dynamic Attention for Multimodal Sarcasm Detection

Dehao Wu and Xingjie Zhuang

Multimodal sarcasm detection has attracted increasing attention due to the widespread use of images and text on social media platforms. Existing studies generally treat the semantic or emotional inconsistency between images and text as a key cue for identifying sarcastic expressions. However, images in real-world scenarios often contain multiple visual objects, many of which are irrelevant to sarcasm-related semantics. Such redundant visual information may introduce noise during cross-modal interaction, distracting models from focusing on critical sarcasm cues and potentially leading to inaccurate predictions. To address this issue, we propose GLDA (Global-to-Local Dynamic Attention), a multimodal sarcasm detection framework that models cross-modal interactions from a global-to-local perspective. Specifically, GLDA first employs a global multimodal fusion module to capture high-level semantic interactions between visual and textual modalities, enabling the model to identify coarse-grained cross-modal inconsistencies. To further reduce the influence of irrelevant objects, GLDA introduces a local gated cross-modal attention mechanism that dynamically filters redundant visual and textual features and refines multimodal representations at a finer granularity. In addition, a multi-branch learning objective is designed to jointly optimize unimodal and multimodal representations. Experimental results demonstrate that GLDA effectively improves sarcasm detection performance and achieves competitive results compared with existing approaches. The code will be available at https://anonymous.4open.science/r/GLDA-DBE6.
Read the full abstract

Computer Vision

From Token Correction to Claim Admission: Training-Free Hallucination Mitigation for Large Vision-Language Models

Hyeonggeun Han, Taehyun Cho and Sangwoo Hong

Large vision-language models often exhibit object hallucination, generating claims that are not supported by the input image. To mitigate this issue, we propose Evidence-Guided Visual Claim Admission (EGVCA), a training-free framework that admits a semantic claim only when it receives sufficient image-dependent support. EGVCA measures this support by comparing the claim likelihood under the observed image and a visual-null input. For captioning, it controls semantic expansion without altering token generation. For binary visual question answering, it combines atomic visual support with the image-induced change in the model's native Yes–No margin. Experiments on CHAIR, POPE, and MME across three LVLM backbones show that EGVCA substantially reduces object hallucination while preserving caption quality. It achieves the highest POPE F1 in eight of nine backbone–sampling settings and improves the native MME total score on all three backbones. These results demonstrate the effectiveness of claim-level admission as a simple alternative to token-level distribution correction.
Read the full abstract

Computer Vision

A Differentiable 3D Scene Graph Metric via Contextual Hellinger Triplet Geometry

U-Chae Jun, Yoonhyuk Choi and Jiwoo Kang

3D scene graph prediction is commonly supervised with local object and predicate classification losses. Although effective for slot-wise label prediction, this formulation does not provide a graph-level semantic distance between a predicted scene graph and its ground truth: semantically mild label errors and structurally disruptive relational errors can be treated similarly, and subject-predicate-object facts are not compared as scene-level units. We propose Contextual Hellinger Triplet Geometry, a graph-level supervision objective that measures the semantic and structural discrepancy between 3D scene graphs. Our key idea is to reinterpret the conventional classifier outputs of a 3D scene graph predictor as an aligned probabilistic scene graph field, where object and predicate predictions jointly define contextual subject-predicate-object facts. This formulation yields a differentiable metric-based training objective with an efficient factorized implementation and can be applied to existing 3D scene graph prediction architectures without modifying their network design. Experiments on 3DSSG demonstrate that our supervision improves downstream graph-conditioned 3D scene generation, while ablation studies confirm the complementary roles of the proposed graph-level terms and user studies show better graph descriptiveness, relation plausibility, and structural preservation.
Read the full abstract

Number Theory

Yau's number-theoretic conjecture and sharp spectral bounds for quantum harmonic oscillators

Zhiwen Liu, Stephen S.-T. Yau and Hao Zuo

We prove Yau's number-theoretic conjecture on positive integral points in real right simplices in every dimension $n\ge3$, and apply the resulting sharp polynomial bound to quantum harmonic oscillators. The lattice-point bound holds for every choice of positive real intercepts for which the count is nonzero, with equality precisely when all intercepts equal the same integer. For oscillators in dimensions $n\ge3$, we obtain explicit upper bounds for eigenvalue counts and spectral remainders for arbitrary positive frequencies at every threshold at or above the lowest eigenvalue. For frequencies with rational ratios in dimensions $n\ge2$, an additional lattice coordinate and interpolation give bounds for energy sums, or first Riesz means, at every threshold. When all frequencies are equal (the isotropic case), the oscillator attains these bounds: the counting bound is exact at its energy levels, and interpolation on its natural energy grid recovers the energy sum at every threshold. In three dimensions, the exact formula displays the periodic contribution associated with the resonance discussed by Fefferman and Seco. The proof builds on the slicing method of Yau and Zhang. Arithmetic and geometric consequences include explicit bounds for smooth numbers and lattice-point discrepancies.
Read the full abstract

Computation and Language

Pushback in a Haystack: Language Models Respond to Factual Errors Ambiguously and Ineffectively

Li Lucy, Myra Cheng, Svenja Guhr et al. (+7 more)

The sycophancy of language models, or their failure to push back against faulty user input, is now widely studied. However, if LMs do push back, is sycophancy solved? We show that LMs vary in how they present pushback, and thus vary in how effectively they correct user beliefs. We introduce a phrase-level, human-validated response labeling framework to uncover fine-grained insights into how LMs push against and/or affirm false inputs. We develop and apply our framework on model responses in five languages: English, German, Chinese, Hindi, and Yoruba. We find that when models do push back against users' false assertions, they tend to use indirect phrasing and mix in affirmation, with more indirectness in Hindi and Yoruba than in other languages considered. In a large-scale human study (n = 300), we find that the pushback ambiguity we identified may fail to correct and prevent users' factual misconceptions.
Read the full abstract

All authors Li Lucy, Myra Cheng, Svenja Guhr, Kabir Ahuja, Orevaoghene Ahia, Sameep Chattopadhyay, Yike Wang, Valentin Hofmann, Noah A. Smith, Yulia Tsvetkov

Artificial Intelligence

A Node Embedding Framework for Model-Centric Conformance Checking in Process Mining

Mostafa Abbasi and Homayoun Najjaran

Evaluating the quality of process models discovered from event logs is a fundamental challenge in process mining. Traditional evaluation relies on conformance metrics such as fitness and precision, which are highly sensitive to noise and deviations in event logs and therefore often fail to give a consistent assessment of discovery algorithms. We introduce a model-centric evaluation framework that combines node embeddings with Procrustes alignment to enable cross-graph comparability, since embeddings derived from different graphs reside in distinct vector spaces. The proposed PetriNetGraph method transforms Petri net process models into graph representations compatible with embedding algorithms. Comparing the node embeddings of ground-truth and discovered models through a parameterized cosine distance yields a unified and bounded metric for evaluating process discovery across datasets and methods. We evaluate the framework under static and context-aware scenarios to examine how prior knowledge of activity existence influences the assessment of structural similarity. The framework also integrates a compliance checking mechanism that localizes deviations at the activity level and an entropy-regularized node distance aggregation function that quantifies global deviations across the model. Experiments on nine publicly available event logs show that the approach provides a reliable embedding-based conformance indicator that reflects the behavior of fitness and precision. Experiments on synthetic logs with controlled noise levels confirm that it is substantially more robust to log deviations than traditional conformance metrics, PetriNet2Vec, and its dynamic-window and Procrustes-aligned variants, while remaining sensitive to process model degradation. All implementation details and datasets are publicly available.
Read the full abstract

Superconductivity

Observation of stacked vortex structure with number controllable Majorana zero mode in Bi$_{2}$Te$_{3}$/FeTe heterojunction

Xinru Han, Zhaokun Luo, Quanxin Hu et al. (+16 more)

Creation and control of Majorana zero modes (MZMs) in solid devices is the key challenge in realizing fault tolerant topological quantum computation. Here we report spectroscopic evidence for number controllable MZMs in stoichiometric one quintuple layer Bi$_{2}$Te$_{3}$/FeTe heterostructures, together with the observation of stacked vortices in superconducting islands. Low temperature scanning tunnelling microscopy and spectroscopy reveal a long period moire superlattice arising from lattice mismatch, which induces nanoscale modulation of superconducting gaps. Magnetic vortices exhibit a stacked core structure with two layer derived bound state branches: a dispersive branch evolving into an X-shaped pattern away from the vortex center, and a quantum-limit branch characterized by a spatially unsplit MZMs together with discrete integer quantized Caroli-de Gennes-Matricon states. Notably, MZMs are observed in all measured vortices, and their number can be tuned by magnetic field and the size of Bi$_{2}$Te$_{3}$ islands. These results establish stoichiometric 1QL-Bi$_{2}$Te$_{3}$/FeTe as a patterned superconducting interface that enables isolated and few-vortex Majorana configurations, providing a controlled platform for fuse MZMs.
Read the full abstract

All authors Xinru Han, Zhaokun Luo, Quanxin Hu, Xingjian Zhu, Xiaohong Pan, Dan Mu, Xinyu Liang, Zulin Xu, Qiheng Wang, Ryan L. Kong, Mingyuan Li, Zhengliang Zhang, Fazhi Yang, Mojun Pan, Xin Liu, Qiangqiang Gu, Baiqing Lv, Gan Wang, Hong Ding

Numerical Analysis

Semi-Lagrangian Approximation of Master Equations with Individual Noise on Finite Graphs

Jianbo Cui and Tonghe Dang

We develop and analyze a fully discrete semi-Lagrangian method for vector master equations associated with mean field games (MFGs) with individual noise on finite graphs. The method discretizes both the characteristic evolution in time and the population variable on the probability simplex. A central challenge is the degeneracy of the transport coefficients near the boundary of the open simplex, where no boundary condition is prescribed. To overcome this difficulty in the analysis, we introduce an interior localization based on exact density characteristics, which keeps numerical arrivals within compact regions of uniform coefficient control. Combined with a closure argument for the coupled numerical recursion, this localization yields local solvability of the updates and quantitative error bounds on compact subsets of the open simplex. Under additional $C^2$-regularity of the exact solution and suitable space–time mesh coupling, these bounds imply first-order accuracy. Furthermore, the density and value trajectories generated by the numerical master feedback converge to the corresponding MFG characteristics at the same rate. In the potential setting, the associated reconstructed density–flux pair is admissible and asymptotically optimal for the planning problem, with an action gap quadratic in the feedback error. Numerical experiments illustrate the convergence rates, feedback responses to population redistribution, and the effects of graph topology.
Read the full abstract

Artificial Intelligence

Understanding Gene Expression-to-Image Learning: Spatial Structure and Information Fidelity

Xiaoqing Lian, Pei Liu, Jiahui Tan et al. (+3 more)

Single-cell RNA sequencing (scRNA-seq) enables cell-level analysis of biological heterogeneity, but learning effective representations from high-dimensional, unordered gene expression profiles remains challenging. Gene Expression-to-Image (GE2I) offers an effective learning strategy by arranging gene values on pixel grids for convolutional neural networks, but its underlying mechanisms remain unclear. We study GE2I on full-transcriptome data while keeping the network and training procedure fixed. Our results reveal a trade-off between spatial smoothness and gene-level fidelity. Structured mappings produce smooth, low-frequency patterns that support faster early optimization, whereas assigning multiple genes to the same pixel obscures expression differences and can reduce predictive performance. This trade-off explains why random mappings can compete with structured layouts when many genes are represented. One-to-one gene-to-pixel assignment improves prediction across the tested visual backbones even when image smoothness decreases. We further find that the benefits of pretraining and augmentation depend on the backbone and mapping. These findings provide practical guidance for pixel allocation and the transfer of vision priors to gene expression data.
Read the full abstract

All authors Xiaoqing Lian, Pei Liu, Jiahui Tan, Quan Zou, Chen Lin, Xiangxiang Zeng

Machine Learning

LeanPRM: Lean-Grounded Process Supervision for Natural Language Proof Verification

Rodney Lafuente-Mercado, Salman Rahman, Jingyan Shen et al. (+3 more)

AI systems now write long and difficult mathematical proofs, and checking their step-by-step validity is becoming a bottleneck. A learned verifier could check proofs cheaply, but training one requires a correctness label on every step, and such labels are hard to obtain for open-ended proofs. Existing methods take them from human annotators, which is costly, or derive them from final answers, which proofs do not have, and formalizing each proof in Lean requires expert effort. We introduce LeanPRM, a framework that uses the Lean proof assistant to create step-level labels for natural-language proofs. A model writes a natural-language proof together with a paired formal proof, without feedback from Lean. Lean then labels every step, and a second model discards failures caused only by the formal syntax, so that the negative labels mark errors in the mathematics. LeanPRM contains 10,634 proofs with 152K labeled steps and no human annotation. A Qwen3-4B verifier trained on LeanPRM raises exact step-label accuracy on a held-out set of 200 proofs (LeanPRM-Bench) from 48.2 to 74.7, as opposed to 60.3 when the same base model is trained on PRM800K. It also selects better proofs in best-of-N selection on IMO-level problems, and outperforms the PRM800K-trained verifier on three of four external verification benchmarks. Training on LeanPRM improves verifiers at every size from 0.6B to 8B parameters on LeanPRM-Bench, Hard2Verify and IMO-GradingBench. Given the labels need no human annotation, LeanPRM can grow with the supply of formal statements, offering a path to step-level feedback on AI-generated proofs that are hard for humans to check.
Read the full abstract

All authors Rodney Lafuente-Mercado, Salman Rahman, Jingyan Shen, James Shiffer, Saadia Gabriel, Pavel Izmailov

Machine Learning

zPocket: Better Pocket Finding with Frozen Co-Folding Features

Tally Portnoi, Allison M Keys, Benjamin R DiFrancesco et al. (+1 more)

Identifying protein-ligand binding sites (pockets) is important for understanding protein function and developing drugs. However, no public method has leveraged co-folding models for novel pocket prediction; furthermore, this task lacks a suitable benchmark as existing benchmarks are within the training distribution of co-folding models. We therefore developed PocketWatch,a curated dataset of biologically relevant ligand pockets. The PocketWatch test set consists of unseen pockets stratified by sequence similarity, structural similarity, and crypticity to enable evaluation of generalization to novel and challenging pockets. Using this dataset, we develop zPocket, a method that trains pocket-prediction heads on top of frozen features from co-folding models. We show that zPocket heads outperform baseline methods across all stratifications. Our best zPocket model achieves overall per-target AUPRC of 0.745 compared to 0.598 and 0.595 for the strongest baselines, VN-EGNN and UniSite-1D retrained on the PocketWatch split (0.559 and 0.531 when using the published weights); the widely used P2Rank reaches 0.477. We further show that zPocket can be used to outperform baseline methods on top-N+2 recall of predicted pocket centroids through clustering per-residue predictions or using zPocket to rescore centroids proposed by other models. Finally, we explore two applications of zPocket: crystallographic fragment screens and a proteome-scale scan for uncharacterized pockets on human proteins. Together these results suggest that zPocket heads built on co-folding trunk features can greatly advance ligand binding site prediction.
Read the full abstract

All authors Tally Portnoi, Allison M Keys, Benjamin R DiFrancesco, Nate Gruver

Optimization and Control

H-PDLP: GPU-Accelerated Linear Programming via Generalized PDHG and Homogeneous Self-Dual Embedding

Haihao Lu and Bo Tang

GPU-based first-order methods are emerging as an approach for solving large-scale linear programming (LP) problems. We present H-PDLP, a GPU first-order LP solver based on a generalized primal-dual hybrid gradient (PDHG) method for a homogeneous self-dual embedding of the LP, and study its numerical performance. The solver applies the restarted reflected Halpern iteration to the generalized PDHG operator, with the embedding encoding both primal-dual optimal solutions and certificates of primal or dual infeasibility. We derive a compact embedding for LPs with two-sided constraint and variable bounds in which the bounds enter through box cones, and we show that the resulting box-cone projections reduce to one-dimensional root-finding problems, so that each iteration remains matrix-free and well suited to GPUs. On the theoretical side, we establish conditions under which the iterates remain uniformly bounded away from the trivial zero solution of the embedding and derive complexity guarantees for the method. We also introduce MIPLIBInfeas, a reproducible benchmark of 229 infeasible and 105 unbounded LPs derived from MIPLIB 2017. Numerical experiments demonstrate the effectiveness of the proposed approach on both feasible LPs and infeasible or unbounded LPs. The source code for H-PDLP is publicly available at https://github.com/MIT-Lu-Lab/H-PDLP.
Read the full abstract

Robotics

HelpMeOut: A Neuropsychology-Grounded Benchmark for Error-Aware Monitoring of Household Task Execution

Haoran Wang, Jingwei Huang, Anqing Duan et al. (+2 more)

We introduce HelpMeOut, an embodied AI benchmark for error-aware monitoring of household task execution. Existing embodied household benchmarks typically assume that humans execute tasks correctly, but this assumption does not hold for cognitively impaired persons or older adults whose everyday activities produce structured execution errors. To address this gap, HelpMeOut operationalizes research on everyday action errors into 20000 paired activity episodes, covering nine neuropsychologically motivated error categories. In each episode, agents observe a streaming human activity trajectory and must monitor whether the execution is unfolding correctly, diagnosing any structured errors as early as possible. We evaluate learned-monitor, symbolic, change-point, and frontier LLM prompting baselines. Results show that these methods perform poorly as error monitors, often missing when execution goes wrong. These results suggest a gap left by standard goal-inference and task-completion benchmarks, making HelpMeOut a testbed for error-aware assistive agents.
Read the full abstract

All authors Haoran Wang, Jingwei Huang, Anqing Duan, Ho Seok Ahn, Bruce Macdonald

Atmospheric and Oceanic Physics

Extreme events and statistics in an AI emulator of a global storm-resolving model

Anna Kwa, W. Andre Perkins, Troy Arcomano et al. (+7 more)

We present HiRO-ACEv2 (High Resolution Output Ai2 Climate Emulator), a two-stage framework for AI emulation of the 3~km resolution X-SHiELD global storm-resolving model (GSRM), and highlight its ability to generate large ensembles of data for studying the characteristics of tropical cyclones and extreme events in high-resolution simulations at an inference computational cost over 300 times lower than that of traditional numerical GSRMs. HiRO-ACEv2 expands upon the precipitation-only HiRO-ACEv1 model to generate coherent precipitation, 10~m winds, and sea level pressure outputs, enabling a broader range of events to be simulated at high-resolution. Fine-resolution predictions are generated by running an autoregressive emulator of coarsened X-SHiELD at 100 km, whose multivariate outputs are regionally downscaled to 3~km resolution via a separately trained diffusion model. Particular importance is placed on capturing the statistics of extreme events during the diffusion downscaling step. Our ability to match the tails of distributions and generate realistic-looking extreme events in both the `perfect prediction' case as well as in free-running emulator rollouts is improved via the use of multiple denoisers at inference time and several updates to the data sampling, noise sampling, and loss weighting during training. HiRO-ACEv2 closely matches the high-resolution reference distributions of precipitation and near-surface winds out to their 99.9999th percentile values, and broadly tracks the reference distribution of tropical cyclone intensities.
Read the full abstract

All authors Anna Kwa, W. Andre Perkins, Troy Arcomano, Noah D. Brenowitz, Christopher S. Bretherton, Spencer K. Clark, Lucas M. Harris, Jeremy McGibbon, Michael S. Pritchard, Oliver Watt-Meyer

Computer Vision

SWORD: Scene-Grounded Wasserstein Object-Relation Distance for 3D Scene Graph Prediction

U-Chae Jun, Jaeeun Ko, Yoonhyuk Choi et al. (+1 more)

3D scene graph prediction is commonly optimized with object-node and relation-edge classification losses. These objectives encourage exact object and relation label recovery, but they do not directly optimize semantic proximity to the ground-truth scene graph: two wrong graphs can receive comparable penalties even when one is much closer in object roles and relation contexts. A direct graph-distance loss is difficult in general, since graph matching and graph edit distance require correspondence search and edit-cost design. Our key observation is that prior supervised 3D scene graph prediction protocols are conducted on fixed object-node indices and supervised edge slots. This can turn a scene graph comparison into a well-defined aligned graph-distance problem. To this end, we propose SWORD, a Scene-Grounded Wasserstein Object-Relation Distance that builds block-normalized object and relation ground distance metrics from training scene graph annotations through a deterministic base-to-context order, without external semantic priors. SWORD lifts these metrics to Wasserstein distances over aligned node and null-augmented edge marginals, providing semantic graph-distance supervision beyond exact-label penalties while preserving the ground-truth graph as the unique zero-distance solution in aligned marginal probability space. We prove that SWORD is a strict metric on aligned marginal probabilistic scene graphs, preserves scene-graph-grounded substitution ordering, controls additive graph-grounded semantic observables, and yields a surrogate upper-bounding decoded semantic error.
Read the full abstract

All authors U-Chae Jun, Jaeeun Ko, Yoonhyuk Choi, Jiwoo Kang

Differential Geometry

Constant Rank Theorems under Intermediate Ricci Curvature

Mijia Lai and Guoqiang Wu

We prove two constant rank theorems for the Ricci tensor in two geometric settings: manifolds with harmonic curvature and gradient Ricci solitons. In both cases, if the $k$-intermediate Ricci curvature satisfies $\Ric_k\geq0$ for some $1\leq k\leq n-2$, then the Ricci tensor has constant rank and its nullity distribution is parallel. In the complete case, the universal cover splits isometrically along the Ricci-null directions. We also show that the range $1\leq k\leq n-2$ is sharp: in every dimension $n\geq4$, nonnegative Ricci curvature alone does not imply the constant rank property in either setting.
Read the full abstract

Artificial Intelligence

MedEvo: Medical Reasoning Evolution through Knowledge Navigation

Xiaoliang Wang, Zhuo Chen, Chuanyang Zheng et al. (+10 more)

Medical multimodal reasoning models are typically trained on teacher-generated rationales, discarding the valid reasoning in students' own failed trajectories, while general-domain correction methods offer little guidance for errors in image reading, anatomical context, or differential diagnosis. We propose MedEvo, which uses medical knowledge, namely a tag taxonomy and error-type diagnosis, to navigate both trajectory repair and data selection. For each failed rollout, a gold-blind auditor locates the first consequential error and provides a minimal, answer-free hint; the same student retains the valid prefix and regenerates the erroneous suffix. After each round, MedEvo identifies capability gaps across medical tags on a held-out feedback set and prioritizes related questions for new rollouts and repairs. With supervised fine-tuning (SFT) alone, MedEvo raises Qwen3.5-9B mean accuracy across seven benchmarks from 65.57% to 69.12% (47.40% to 54.50% on the multimodal track of MedXpertQA (MedXpertQA-MM)), surpassing teacher-distillation SFT (66.93%). It also raises Qwen3.5-35B and Qwen3.5-397B mean accuracy from 67.91% to 73.04% and from 71.83% to 74.95%, respectively. Our analyses show that failed trajectories contain reasoning worth preserving, that repair success does not imply learning value, and that independent feedback is key to sustained improvement.
Read the full abstract

All authors Xiaoliang Wang, Zhuo Chen, Chuanyang Zheng, Quanbin Wang, Jiangwei Lao, Yingao Liu, Jiajia Liu, Jinming Cao, Honglin Lin, Zhihong Lu, Hao Jin, Jian Wang, Lijun Wu

Artificial Intelligence

Stable Token-level Scheduling for Serverless LLM Inference

Hardi Ashishbhai Desai, Balaji N K and Sourav Kanti Addya

Serverless inference is emerging as an efficient deployment model for Large Language Models (LLMs) due to its ability to elastically scale resources, improve GPU utilization through resource sharing, and reduce infrastructure costs. However, existing approaches primarily optimize time to first token (TTFT) while overlooking inter-token latency (ITL), a key determinant of user experience during response generation. We observe that LLM inference comprises a computation-intensive prefill stage and a bandwidth-bound decode stage, where admitting new requests can disrupt ongoing decoding and increase latency. To address this, we propose Token Window Pacing (TWP), a closed-loop scheduling framework that leverages runtime feedback from ITL, queue depth, and workload characteristics to regulate request admission at the token level instead of the batch level. TWP dynamically adjusts a selfregulating token window and incorporates chunk-based prefill to minimize interference from incoming requests. Experiments on realistic serverless workloads show that TWP delivers stable TTFT, significantly reduces latency spikes, lowers P99 tail latency, minimizes ITL variation, and improves throughput by up to 45% over ServerlessLLM and 58% over conventional schedulers. These results demonstrate that feedback-driven tokenlevel scheduling enables stable, high-throughput serverless LLM inference without modifying the underlying model.
Read the full abstract

Robotics

OccWAM: Occupancy World-Action Models with Joint Scene-to-Ego Flow Matching

Qihang Huang, Wenhui Huang, Collister Chua et al. (+6 more)

Occupancy world models provide a structured, metric representation of future driving scenes, making them a promising basis for end-to-end autonomous driving. However, existing occupancy-aware planners typically couple occupancy prediction and action generation through either hierarchical coupling or parallel multi-task learning. Hierarchical methods use occupancy either as a fully predicted future preceding action generation or as an action-conditioned rollout for trajectory evaluation, whereas parallel multi-task methods predict occupancy alongside planning through a separate branch or auxiliary objective. These interfaces retain either an explicit generation order or a role asymmetry, limiting reciprocal interaction between world prediction and action generation during trajectory formation. We propose OccWAM, an occupancy-native world-action model that jointly generates dense future occupancy and continuous ego trajectories via flow matching conditioned on scene representations. Building on this joint formulation, we introduce scene-to-ego world-action flow matching, a two-stage formulation. The first stage predicts scene-centric occupancy together with an absolute ego trajectory, providing a stable representation of future scene evolution. The same predicted world is then re-indexed into the future ego frames induced by the initial trajectory, where ego motion is represented through adjacent pose increments and a second joint flow refines ego-centric occupancy and motion for action-aligned planning. Unlike video-based world-action models, OccWAM preserves explicit semantic geometry for direct planning reasoning without pixel synthesis. Extensive experiments on Bench2Drive and nuScenes demonstrate state-of-the-art performance in closed-loop driving, open-loop planning, and future occupancy prediction.
Read the full abstract

All authors Qihang Huang, Wenhui Huang, Collister Chua, Shiyu Fang, Zhan Chen, Songyan Zhang, Zhidong Wang, Daiying Tian, Chen Lv

Materials Science

A Parameter-Free Reference Hamiltonian for Tight-Binding Vanadium Oxides

Jun-qiang Lu

Density-functional tight binding has yet to yield a transferable description of vanadium oxides; for the standard second-order construction and the tested cells, we demonstrate that two of its approximations deviate by electron volts and offset each other in the band-index centroid separation between the V 3d and O 2p bands: the neutral two-centre reference omits electron-volt environmental terms, whereas the Mulliken charge model assigns the compact V 3d shell less than half of the self-consistent potential shift observed in PBE. Instead, we build a tight-binding model where the reference Hamiltonian derives from superposed neutral atoms, the full PBE exchange-correlation potential and frozen-core orthogonality, without adjustable terms, and where the orbital-resolved charge response employs three screening factors obtained from PBE linear response and an unscreened Coulomb range. Relative to the primary PAW reference, the plane-wave Harris-Foulkes energy differences of VO2 are matched by the reference energy at fixed geometry and the stated numerical settings, with signed deviations of +2.6 meV/atom for M1 relative to rutile (an energy difference of 92 meV/atom) and +6.2 and -5.2 meV/atom under 3% volume compression and expansion; the charge response reproduces 14 of 22 fitted and 5 of 8 held-out matched PBE responses within one tolerance, and 28 of the 30 within twice the tolerance, and in rutile the fitted first-radial V 3d shift is -4.39 eV, 0.47 eV more negative than PBE's -3.92 eV, whereas the tested DFTB2 charge model gives -1.53 eV. The convexity of the spinless model at fixed geometry prevents it from spontaneously differentiating symmetry-equivalent sites under the symmetries respected by its implemented operators and sampling, so within this model such differentiation in the Magnéli phases must arise from electron-lattice coupling or an explicit spin or correlation term.
Read the full abstract

Computation and Language

Encoding Context into LLM Weights: What Makes It Context?

Yue Wang, Qizhou Wang, Feng Liu et al. (+3 more)

Longer contexts allow large language models (LLMs) to condition their predictions on more information, but inference costs grow prohibitively with context length. An emerging alternative uses test-time updates to encode context as temporary changes to model weights. However, encoding this information into weights that already represent general knowledge obscures its role as context, so the LLM may fail to use it properly to condition subsequent inference. To restore this contextual role, we first associate each context with an identifier during encoding, thereby distinguishing the newly encoded information from pre-existing knowledge. Invoking the identifier at inference then enables the LLM to recite the encoded context. Nevertheless, successful recitation does not ensure that the LLM will use the recalled information as context. We therefore further introduce a prompting framework that explicitly guides the LLM to use this information as context. Across diverse controlled tasks and practical benchmarks, our method outperformed the baseline using only off-the-shelf LLMs, showing its advances in context encoding.
Read the full abstract

All authors Yue Wang, Qizhou Wang, Feng Liu, Gang Niu, Masashi Sugiyama, Bo Han

Computation and Language

LatentRiskBench: Measuring State-Dependent Risks in Enterprise Agents

Jingyue Cong, Estrid He, Zhuang Li et al. (+5 more)

Enterprise agents face safety risks that cannot be determined from individual actions alone. Existing safety benchmarks largely evaluate action-local risks, overlooking failures whose safety depends on workflow state. In such state-dependent risks, the same action may be safe in one context but violate policy in another. Hence, we introduce LatentRiskBench for evaluating such state-dependent risks in enterprise settings. It pairs risky trajectories with benign counterparts that share the same terminal action but differ in latent policy-relevant states, requiring models to reconstruct workflow context rather than rely on action-level cues. We also introduce new metrics for state-dependent risk evaluation, including a suite of residual risk (RR) measures. Experiments on the benchmark show that action-level evaluation detects only 0.09 of state-dependent violations, whereas state-level evaluation reaches 0.96. Existing guardrail methods largely overlook such risks unless they are explicitly applied to workflow states. Case studies on a commercial deployed scenario further show that the safety metrics embedded in the benchmark provide a more realistic measure of safety for enterprise agents. Our dataset and code are available at https://github.com/JingyueCong/LatentRiskBench.
Read the full abstract

All authors Jingyue Cong, Estrid He, Zhuang Li, Nan Wu, Li Shen, Yameng Peng, Xinshu Li, Andy Song

Computer Vision

Hierarchical Curriculum Learning for Surgical Video Generation: Focusing on Bilateral Axillo-Breast Approach Robotic Thyroidectomy Datasets

Youngbin Kong, Seungho Lee, Jae Kyung Shin et al. (+3 more)

High-fidelity surgical video generation is increasingly important for surgical education, simulation, and downstream AI training, as conventional simulation pipelines are limited by cost, realism, and scenario diversity. Adapting generic text-to-video diffusion models to surgical procedures remains difficult, however, due to complex anatomical structures, intricate tool-tissue interactions, the inherently hierarchical nature of surgical workflows, and the scarcity of curated training data outside a few well-studied procedures. We propose a surgical video generation framework that combines hierarchical procedural annotation with curriculum transfer learning. Surgical procedures are decomposed into three levels (phase, task, and action) and a pretrained text-to-video diffusion model is progressively adapted from coarse to fine-grained surgical context, supported by a dynamic sampling strategy that handles variable-resolution and variable-length surgical clips. We instantiate the framework on Bilateral Axillo-Breast Approach robotic thyroidectomy, a procedure for which no curated benchmark currently exists. Relative to the preceding task-level model, the final action-level model reduces FVD by 20.1% and FID by 15.7%, and a matched-budget ablation indicates that these gains arise from the curriculum design rather than from additional training. A user study with six clinicians rates visual realism highly while highlighting remaining gaps in anatomical accuracy and clinical reliability. The proposed framework offers a step toward AI-driven surgical simulation in procedures where curated data are scarce.
Read the full abstract

All authors Youngbin Kong, Seungho Lee, Jae Kyung Shin, Jong-hyuk Ahn, Hyoun-Joong Kong, Kyu Eun Lee