Robotics
RESTA: Time-Scaling Augmentation for Data-Efficient Robot Imitation Learning
Yunpeng Jiang, Jianshu Hu, Han Fang et al. (+3 more)
Robot manipulation policies trained through imitation learning commonly rely on costly demonstrations recorded in one task direction and at one temporal resolution. We propose Reversal and Stride-based Trajectory Augmentation (RESTA), which exploits time-scaling data augmentation to improve data efficiency through trajectory reversal and temporal resampling. Trajectory-reversal augmentation reverses and relabels demonstrations between paired tasks with opposite objectives, reusing recorded rollouts without additional environment interaction. Because contact-rich manipulation is not exactly reversible, we treat reversal as an approximate paired-task prior rather than a physical
symmetry. Temporal-stride augmentation varies observation and action sampling strides and conditions the policy on the resulting offsets, enabling fixed or randomized temporal schedules. Experiments across 16 paired bimanual manipulation tasks show that trajectory reversal is particularly effective in the low-data regime and that temporal-stride augmentation improves success under randomized temporal schedules. With 25 and 50 independently collected demonstrations per task, RESTA achieves 78.4% and 85.2% average success, respectively, matching or exceeding baselines trained with twice as many demonstrations (77.3% and 85.4%). These results show that RESTA can improve data efficiency in
robot imitation learning.
Read the full abstract ▾
All authors Yunpeng Jiang, Jianshu Hu, Han Fang, Zhiyuan Guan, Paul Weng, Yutong Ban
Machine Learning
LoOPSD: On-Policy Self-Distillation with Looped Teachers
Wei Fan, Qingyue Yang, Chao-Hong Tan et al. (+6 more)
On-policy self-
distillation improves a model using guidance from the model itself, but it still requires a teacher that provides more useful predictions than the student. Existing approaches obtain such guidance by giving the teacher additional context or instructions. We ask whether it can instead come from changing how the same model uses its own layers. To address this, we introduce LoOPSD, which searches over a model's layer execution order to construct a stronger teacher and distills its advantage into the student. Specifically, we keep the model weights unchanged while searching for layers to repeat, and use the resulting teacher's predictions on student-generated prefixes to guide learning. We then alternate search and
distillation, rebuilding the teacher from the updated student in each round so that the student progressively absorbs the gains from alternative layer execution. Experiments on mathematical and logical
reasoning tasks show improvements across models. On GSM8K with a 400-
token thinking budget, LoOPSD raises Qwen3-1.7B accuracy from 42.30% to 79.61%; subsequent
distillation rounds yield further gains beyond the first round. Together, these findings show that by searching layer execution to construct teachers, we can turn a model's additional computation into improved student capabilities, opening a new avenue for self-
distillation to derive supervision from the model's own computation.
Read the full abstract ▾
All authors Wei Fan, Qingyue Yang, Chao-Hong Tan, Qian Chen, Wen Wang, Xiangang Li, Kejiang Chen, Weiming Zhang, Nenghai Yu
Computer Vision
Benchmarking Multimodal Large Language Models for Zero-Shot Human-Object Interaction
Oriol Rabasseda, Zenjie Li, Kamal Nasrollahi et al. (+1 more)
Multimodal Large
Language Models (MLLMs) have demonstrated strong zero-shot
reasoning across diverse domains, but their potential for Human-Object Interaction (HOI) remains unexplored. Unlike traditional HOI approaches that rely on extensive labeled
datasets and fixed interaction taxonomies, MLLMs offer flexible
reasoning, making them well-suited for dynamic, real-world scenarios. In this work, we present the first systematic study of MLLMs for HOI recognition and detection within a pipeline that combines an external object detector for entity localization with zero-shot
multimodal reasoning for interaction understanding. We introduce a
benchmark framework for principled zero-shot evaluation, consisting of (i) a unified taxonomy of zero-shot prompting strategies, and (ii) a methodology that transforms
token-level logits into interaction scores, enabling fair comparison with specialized HOI models under standard metrics. To support
prompt-based evaluation, we additionally construct new question-answer annotations for standard HOI
datasets, transforming them into a format suitable for MLLMs. Building on this, we propose two visual prompting strategies, namely Objects as Context and Black Other Objects, that guide MLLMs to target objects while preserving scene context. Experiments on HICO, HICO-DET, and V-COCO demonstrate state-of-the-art zero-shot performance, achieving 53.50 mAP on HICO and 23.69 mAP on HICO-DET. Overall, our work provides a principled benchmarking framework for MLLMs in HOI, opening the door to new models and
datasets and enabling flexible open-world interaction understanding.
Read the full abstract ▾
All authors Oriol Rabasseda, Zenjie Li, Kamal Nasrollahi, Sergio Escalera
Robotics
Force Sensor-Based Picking Pattern Analysis and Two-Jaw Gripper Design for Cherry Tomato Robotic Harvesting in Greenhouse
Rajmeet Singh, Manveen Kaur, Yusuf Adamu et al. (+1 more)
Robotic harvesting of cherry tomatoes in greenhouses remains challenging due to dense fruit clusters, delicate skin, and the need for gentle yet effective detachment. This study presents a force-data-driven framework for autonomous cherry tomato harvesting that integrates six-axis force/torque sensing, a bio-inspired compliant gripper, and deployment on a mobile robotic platform. A wearable ATI six-axis force/torque sensor mounted on the index finger captured kinetic signatures of three elemental picking patterns (gripping, twisting, and lateral pulling) across 150 manual trials (50 per pattern) in laboratory and greenhouse settings. Gripping required a peak normal force of 3.92 and 0.45 N, twisting produced a dominant torque of 19.20 2.10 Nmm about the pedicel axis with minimal truss disturbance, and pure lateral pulling alone was insufficient for reliable detachment. These measurements informed a novel fingertip-localised Finray gripper, with compliant TPU 95A Finray geometry restricted to the fruit-contact region and a rigid PLA body, emulating the dual-stiffness human fingertip. A Dynamixel XM430-W350 servomotor under current-based torque control applies the measured gripping force and a 45{\deg} wrist twist. Finite element analysis showed a maximum von Mises stress of 0.24 MPa under the 3.92 N load, below the bruising threshold of cherry tomato skin. Integrated on a UR5 arm and Clearpath Husky A200 base, the system was validated in NVIDIA Isaac Sim and in 50 autonomous trials at the SILAL commercial greenhouse, Abu Dhabi, UAE. The hybrid grip-twist-pull strategy achieved an 86.0% picking success rate, 2.3% fruit damage, and an 8.4 s cycle time per fruit, outperforming
vacuum (66.3%), rotating (70.1%), clamping (76.0%), and suction-cutting (52.4%) end-effectors. These results demonstrate preliminary feasibility under the tested facility and cultivar conditions
Read the full abstract ▾
All authors Rajmeet Singh, Manveen Kaur, Yusuf Adamu, Irfan Hussain
Computation and Language
Claim Framing as a Driver of LLM Sycophancy
Jean Seo, Pepa Atanasova and Isabelle Augenstein
Sycophancy in Large
Language Models, the tendency to align responses with user preferences over ground truth, has largely been studied using questions without a clear ground truth and with a single, plainly stated user opinion. We present a systematic study of which framings of a user's claim trigger sycophancy in factual question answering, where every answer is objectively verifiable without extra context. Evaluating 14 models on 42K questions, we express correct and incorrect user opinions through various claim framings over 8 dimensions and 21 types, and measure a broad set of opinion-driven behaviors. We find that all models are susceptible to user opinions regardless of size. Models are most influenced by how confidently an opinion is stated, whom it is attributed to, and whether it is presented as shared between the user and the model. In contrast to prior work, framings based on the user's expertise, authority over the model, or social pressure matter little in our factual setting. Larger models resist incorrect opinions better but are no better at recovering from their own errors. Our results call for sycophancy evaluations that probe how an opinion is framed, not merely whether one is present, and that measure multiple behaviors beyond a single flip rate.
Read the full abstract ▾
Artificial Intelligence
When and How Order-Flow Interventions Work After Dealers Adapt
Guanghao Wu, Zhuo Cai, Zhidong Li et al. (+1 more)
Learning dealer
agents can settle on persistently wide quotes, raising what customers pay to trade. A venue can route more orders to tighter quotes, but a wide quote may reflect real risk, and once dealers adapt, an end-to-end comparison cannot show why a rule works: because it intervenes more, picks better moments, is announced in advance, or rewards the right quote. We propose SOFA (State-aligned Order-Flow Allocation), a public order-flow contract with three parts: a State Scorer ranks
market states from lagged public data, a Commitment Scheduler announces a budgeted intervention strength before dealers quote, and a Quote-to-Flow Allocator shifts a bounded share of fixed customer demand toward tighter quotes. Our paired re-adaptation protocol gives competing rules the same intervention strengths, only re-timed, lets dealers re-adapt to each rule from a shared checkpoint, and breaks one link at a time. Across 20 independently trained dealer populations in an ABIDES
market, concentrating the intervention in states that the score ranks as low-risk lowers low-risk execution cost relative to random timing in every population, by 0.040 tick on average, while a stale announcement or a broken quote-to-flow link raises this cost. In the primary setting the gain comes at the expense of high-risk trades, whose costs rise, so state
alignment pays off only for a venue that weighs low-risk costs more than 1.38 times as heavily as high-risk costs. Our code is available at https://github.com/moderncavemann/state-aligned-order-flow-contracts.
Read the full abstract ▾
All authors Guanghao Wu, Zhuo Cai, Zhidong Li, Shoujin Wang
Computation and Language
Fan Out or Relay? Ownership-Aware Routing for LLM Agents Querying Siloed Tables
Zhexi Feng
Inside organizations, a question's data often sit with several owners, who query their own tables but never hand them over. An LLM coordinator can fan out, sending the question to every owner at once, or relay, querying owners in turn and reusing earlier rows. We show that neither is safe in advance, because data ownership decides which one is right. Across thirteen settings on PeopleJoin and BIRD with three model families, fan-out is more accurate in twelve on questions one owner can answer, yet answers at most 4.3% of those joining two owners' tables. The flip follows ownership, not query shape: on BIRD two-table joins, fan-out answers 32-44% when one owner holds both and 1-4% otherwise. Relay wins by conditioning on returned rows, and hiding them from an otherwise identical planner cuts two-owner accuracy from 23.2% to 5.1%. Since ownership is public, we propose Ownership-Aware Routing (OAR), which drafts a schema-only SQL query, counts the owners whose tables it joins, and routes the question before any private row is read. In ten settings OAR is never significantly less accurate than the best fixed protocol, and beats every fixed protocol on PeopleJoin, by up to 9.3 points on held-out organizations.
Read the full abstract ▾
Optimization and Control
UQ-ACE: Stability-Aware Counterfactual Controller Tuning via Bayesian SINDy
Margarita A. Guerrero, Rodrigo A. Gonzalez and Cristian R. Rojas
Retuning feedback controllers in critical processes is challenging when experiments are costly, open-loop excitation is infeasible, and candidate controllers must be screened before deployment. The existing Adaptive Sampling for Counterfactual Explanations (ACE) method provides a sample-efficient approach for finding minimal controller-parameter changes that recover prescribed closed-loop performance, but it does not explicitly account for stability of candidate controllers proposed during the search, which may be destabilizing under plant uncertainty. This paper introduces UQ-ACE, a stability-aware retuning method based on
Bayesian SINDy. In UQ-ACE, closed-loop trajectories are used to infer a posterior distribution over local dynamical models, from which candidate controllers are screened for stability before implementation. Extensive simulations show that UQ-ACE achieves valid retunings with few closed-loop experiments while preventing deployment of unstable candidates.
Read the full abstract ▾
Machine Learning
When Do Learned Blocks Preserve Causal Meaning? Block Identifiability and Constructive Causal Abstractions
Karim Zaghw, Wieland Brendel and Patrik Reizinger
Standard identifiability theory assumes one-dimensional latent factors, but learned representations can encode latent information across multidimensional blocks whose internal coordinates are not uniquely determined. When such blocks are used to represent and intervene on higher-level concepts, it is unclear whether their causal meaning survives the transformations left unresolved by block-identifiability. We connect block-identifiability to constructive causal abstraction. We show that block-diagonal invertible transformations followed by compatible coordinate permutations form the maximal class that preserves every constructive causal abstraction built on those blocks. This guarantee assumes that the correspondence between blocks and concepts is updated with the representation. A small synthetic experiment with additive autoencoders shows that high-level interventions can be transferred between independently learned representations while approximately preserving their intended effects.
Read the full abstract ▾
Machine Learning
Learning a Group Element from Its Action: Galois Correspondences, Optimal ERM, and an Abel-Ruffini Threshold for the Natarajan Dimension
Xiaoyu Li
We study learning an unknown element $g$ of a group acting with finite orbits on a countable set, from examples $(x,gx)$. After a sample on $A$, the consistent hypotheses are the restrictions of a coset of the pointwise stabilizer of $A$, and the sample forces exactly the labels of points this stabilizer fixes; for a Galois group, they form the field generated by the data. Through this correspondence, multiclass dimensions become group invariants: the DS and graph dimensions equal the height, the Littlestone dimension the longest irredundant sequence length, and the teaching dimension the minimal base size. Among deterministic learners,
symmetry makes empirical risk minimization optimal: on nondegenerate classes every consistent learner, whatever its tie-breaking, attains the optimal realizable sample complexity up to constant factors, and every consistent online learner the optimal mistake bound. Agnostically, every ERM's sample complexity exceeds the known lower bound by at most the DS-to-Natarajan ratio, up to constants. For finite Galois extensions of countable fields, the Natarajan dimension equals a group invariant, the coset-cube number, which is at least the number of complemented factors of any chief series of any subgroup. Hence the dimensions coincide for every solvable Galois group, while for the groups $A_5$ and $S_5$ of the quintic they are $2<3$ and $3<4$, and $A_5$ is the smallest group with a gap. For the Galois group $S_n$ of the general equation of degree $n$, they coincide exactly when $n\le 4$, i.e., in characteristic zero, exactly when the equation is solvable by radicals; the implication is one-way, as the simple group $\mathrm{PSL}(2,8)$ has no gap. For linear, orthogonal and
unitary maps of $\mathbb R^d$ or $\mathbb C^d$ both dimensions equal $d$, and a two-fibered pseudo-cube is a disjoint union of Natarajan cubes exactly when its monodromy group is abelian.
Read the full abstract ▾
Optics
Experimental Demonstration of 7-Dimensional N-Vector Model with Dimensional Inflation and Annealing
Daniele Veraldi, Marcello Calvanese Strinati, Paolo Charlesworth et al. (+4 more)
We report the first experimental realization of a physical system described by a high-dimensional
Hamiltonian comprising up to seven dimensions. Our setup is based on a holographic optical computer, which realizes a dimensional inflation by spatial and temporal multiplexing of lower dimensional systems. We demonstrate an $N$-vector model, with applications in field theory and computing. The approach is scalable and enables table-top implementation of high-dimensional theoretical models, and novel methods for combinatorial optimization. We experimentally demonstrate dimensional annealing: by dynamically reducing the number of dimensions, the ground state of an Ising model is found in complex graphs by an adiabatic
evolution from high-dimensional spaces.
Read the full abstract ▾
All authors Daniele Veraldi, Marcello Calvanese Strinati, Paolo Charlesworth, Matteo Spaziani, Silvia Gentilini, Davide Pierangeli, Claudio Conti
Combinatorics
Alternating Hamilton Paths and Cycles in Poset Products
Yuxian Dong and Jianxi Mao
We study alternating Hamilton paths and cycles in the comparability graphs of finite posets. We prove that if a finite poset $P$ with at least two elements admits an alternating Hamilton path, then $P\times B_r$ admits an alternating Hamilton cycle for every $r\ge2$, where $B_r$ is the Boolean
lattice. As a corollary, we settle a
conjecture of Katona [Discrete Appl. Math., 2013]. Furthermore, we characterize the divisor
lattices admitting alternating Hamilton paths and cycles.
Read the full abstract ▾
Computation and Language
Attacking Only Half the Data: Two-Sided Evaluation of AI-Text Detector Robustness
Apu Chandraw Shill and Rabindra Nath Nandi
We investigate the robustness of widely used AI text detectors under attacks applied to both human and machine text. Prior evaluations typically attack only machine text, overlooking that human text is routinely modified through grammar correction, formatting, or invisible character insertion. We introduce a two-sided protocol that applies attacks to both classes. Across two corpora and nine detectors, reported AUROC gains vanish under two-sided evaluation, with attacked human text misclassified at rates up to 100%. While simple preprocessing reverses character-level attacks, word-level synonym substitution remains effective meaning-preserving, and irreversible.
Read the full abstract ▾
Functional Analysis
Reverse Translation and Convolution Estimates in the Framework of the Mehler-Fock Transform
U. K. Mandal, Akhilesh Prasad and Sudhanshu Ranjan
In this paper, reverse translation and convolution estimates in the frame of $\mu^{th}$ order Mehler-Fock transform are obtained. Continuity properties of these operators on the test function space $\mathfrak{H}_{l,r,n}$ have been discussed. Moreover, Sobolev space $S_{r,p}^{N,\alpha}$ has been introduced and again continuity properties of these operators in this Sobolev space have been discussed.
Read the full abstract ▾
Rings and Algebras
Standard identities for polynomial vector fields
Nurtas Shyntas
Let $W_n$ be the Lie
algebra of
polynomial vector fields in $n$ variables over a field of characteristic zero, and let $s_k$ denote the alternating sum of compositions of $k$ vector fields. The identity $s_{n^2+2n-1}=0$ has been asserted for arbitrary $n$. We give an explicit counterexample in dimension five, verified by an exact integer computation of a $34$-commutator. Together with the known vanishing of $s_{35}$, this shows that $35$ is the least degree of a standard identity on $W_5$. We then prove that $s_{n^2+2n-1}=0$ for every even $n\geq2$. The proof uses Lie
algebra homology with coefficients in the
polynomial algebra modulo constants and a lemma on invariant tensors. We describe the computation and provide the source code and verification results.
Read the full abstract ▾
Dynamical Systems
Topological entropy of saturated sets for Lorenz attractors
Zheng Huang, Xueting Tian, Xiaodong Wang et al. (+1 more)
We study the topological
entropy of saturated sets for continuous flows, with emphasis on singular
hyperbolic dynamics. For a $C^r$-generic ($r \geq 2$) vector field $X$ carrying a
geometric Lorenz attractor $\Lambda$, we prove that for every nonempty compact connected set $\mathcal{K} \subseteq \mathcal{M}(\Phi,\Lambda)$ and every open set $U \subseteq M$ with $U \cap \Lambda \neq \emptyset$, $h_{\mathrm{top}}(\Phi,G_{\mathcal{K}}(\Phi) \cap \operatorname{Trans}(\Phi,\Lambda) \cap U)=\inf_{\mu \in \mathcal{K}} h_\mu(\phi_1)$, where $\Phi=(\phi_t)_{t \in \mathbb{R}}$ is the flow induced by $X$ and $\operatorname{Trans}(\Phi,\Lambda)$ denotes the set of points whose positive orbits are dense in $\Lambda$. The same formula holds $C^1$-generically for singular
hyperbolic attractors. In contrast with the uniformly
hyperbolic case, we also prove that, for vector fields in a dense subset of $\mathcal{X}^r(M^3)$,
geometric Lorenz attractors admit a positive-
entropy invariant measure with no generic points. An analogous conclusion holds for vector fields in a dense subset of $\mathcal{X}^1(M)$ whose non-trivial singular
hyperbolic attractors contain singularities and satisfy $\dim E^{cu}=2$. These examples show that the discrete-time saturated
entropy theory does not extend directly to singular flows.
Read the full abstract ▾
All authors Zheng Huang, Xueting Tian, Xiaodong Wang, Xutong Zhao
Robotics
Walk Like This: Stylized Locomotion for Physics-Based Characters through Example-Driven Diffusion
Joshua Aurand, Agon Serifi, Sammy Christen et al. (+4 more)
We present a real-time framework for expressive legged locomotion that enables independent control over velocity and style. Unlike existing data-driven controllers, which typically require explicit style labels for training, our method efficiently leverages large-scale unstructured motion data and generalizes to unseen styles. We disentangle style from functional locomotion through a tailored sampling strategy used to train a style encoder whose output conditions an autoregressive
diffusion model. During inference, the user provides a style reference clip and a joystick velocity command. The
diffusion model then generates a kinematic reference tracked by a reinforcement learning controller. To improve user controllability and stylization strength, we propose an
attention biasing mechanism and a velocity command correction strategy. We demonstrate our method on a physics-based humanoid character in simulation and on a physical
robot, showcasing robust real-time control and continuous style modulation. The results can be best seen in our supplemental
video (https://www.youtube.com/watch?v=roFnS6N7nJ4).
Read the full abstract ▾
All authors Joshua Aurand, Agon Serifi, Sammy Christen, Ruben Grandia, Espen Knoop, Marco Hutter, Moritz Bächer
High Energy Astrophysical Phenomena
Low-frequency spectra and scattering of pulsars towards the Gum Nebula with the MWA
Ting Yu, Zhifu Gao, Zhongli Zhang et al. (+1 more)
Context. The Gum Nebula is a nearby ionised complex with sight-line-dependent
pulsar propagation. Aims. We test whether a smooth ionised structure can explain low-frequency spectral turnovers and pulse broadening, or whether smaller-scale material is required. Methods. We analysed Murchison Widefield Array observations of six
pulsars at 80–240 MHz, combining sub-band flux densities with literature data. We compared power-law, free–free absorption (FFA), and turnover-constrained broken-power-law (BPL) spectra, and modelled scattering through pulse-profile convolution. PSR J0820-4114 and the Vela
pulsar were fitted jointly across frequency. Results. All six spectra depart from a single power law. FFA is favoured for PSRs J0737-3039A and J0820-4114, and BPL for J0742-2822 and J0835-4510; the other two classifications depend on assumed error floors. Conditional FFA solutions imply emission measures of $10^4$–$10^5$ pc cm$^{-6}$ for electron temperatures of $3\times10^3$–$10^4$ K, far exceeding smooth-shell expectations. PSRs J0737-3039A and J0837-4135 have scattering laws consistent with $\tau_{\rm sc}\propto\nu^{-4}$. We obtain $\alpha_{\rm sc}=-3.07\pm0.11$ for J0742-2822 and $-1.59\pm0.44$ for J0820-4114 using a joint two-component model. The Vela
pulsar is best described by an effective thick-screen model with $\alpha_{\rm sc}\simeq-2.74$. Neither tested Galactic electron-density model reproduces both independent distances and low-frequency scattering uniformly. Conclusions. The results favour a structured Gum–Vela magneto-ionic environment in which neighbouring sight lines probe different absorbing and scattering material; these effects need not arise in the same
plasma component.
Read the full abstract ▾
All authors Ting Yu, Zhifu Gao, Zhongli Zhang, Weicong Jing
Atmospheric and Oceanic Physics
Linking Horizontal and Vertical Submesoscale Lagrangian Dynamics South of the Ibiza Channel
Sara Cloux, Ismael Hernandez-Carrasco, Alejandro Orfila et al. (+2 more)
Fronts, eddies, jets, and meanders populate the ocean surface and play a key role in organizing transport and mixing across a wide range of scales. From a Lagrangian perspective, frontal structures are conveniently detected as surface filaments that act as barriers to horizontal transport. However, the degree to which these structures maintain coherence below the surface and the relationship between horizontal and vertical motions in their vicinity remain poorly understood. Here we analyze two such filaments in the western Mediterranean Sea, south of the Ibiza Channel, using a high-resolution ocean simulation and Lagrangian diagnostics specifically designed to characterize submesoscale transport. Backward Finite-Size Lyapunov Exponents (FSLE) computed at successive depths reveal nearly vertical, curtain-like structures extending several tens of meters into the water column, with intensity decreasing with depth. Vertical-velocity sections across the filaments show coherent downward motion aligned with the filament cores, while Lagrangian trajectories indicate rapid horizontal convergence of fluid particles accompanied by sustained downward displacement. The barrier effect of Lagrangian Coherent Structures (LCSs) is commonly associated with persistent dynamical features of large-scale flows. Here, however, we demonstrate that FSLE filaments generated by rapidly fluctuating submesoscale dynamics can also act as effective transport barriers. The results demonstrate that the surface signatures of Lagrangian filaments identified through FSLE techniques correspond to three-dimensional transport structures that regulate both lateral aggregation and vertical connectivity in the upper ocean.
Read the full abstract ▾
All authors Sara Cloux, Ismael Hernandez-Carrasco, Alejandro Orfila, Emilio Hernandez-Garcia, Cristobal Lopez
Data Structures and Algorithms
A Better-than-2 Approximation for k-Median Part I: The Power of General Bids
Rares-Darius Buhai, Vincent Cohen-Addad, Henry Fleischmann et al. (+6 more)
The Greedy
algorithm of Jain, Mahdian, Markakis, Saberi, and Vazirani was the best Lagrangian-multiplier-preserving (LMP) approximation
algorithm for facility location, and thus the engine behind the best $k$-median guarantees, for over two decades. Its $2$-LMP guarantee was only recently broken by a small amount ($(2-\eta)$-LMP where $\eta \approx 2.25 \cdot 10^{-7}$) through local-search post-processing. That analysis relies on the structure of worst-case instances, and it appears difficult to obtain substantial further improvements along those lines. We take a different approach and change the
algorithm itself: we introduce general bidding functions, replacing the classical linear bid with a richer family of rules and prove a
polynomial-time $1.95$-LMP approximation, a large improvement over the long-standing factor of~$2$. We further show that these generalized bidding rules can be made smooth (log-adaptive), building on the recent framework developed for the classical Greedy
algorithm by Cohen-Addad, Grandoni, Lee, Schwiegelshohn, and Svensson. This yields a
polynomial-time $1.95$-approximation that opens $k+O_\varepsilon(\log n)$ centers. Combined with a $(1.89+\varepsilon)$-approximation
algorithm for stable instances in our companion work \cite{stableinstances}, we achieve a better-than-2 approximation for the $k$-median problem in
polynomial-time.
Read the full abstract ▾
All authors Rares-Darius Buhai, Vincent Cohen-Addad, Henry Fleischmann, Anupam Gupta, Antoni Jubés Monforte, Madhusudhan Reddy Pittu, Miltiadis Stouras, Ola Svensson, Lukas Vogl
Artificial Intelligence
RadixFlow: Filling Execution Bubbles in Prefix-Cached Hybrid LLM Serving
Huihao Jing, Wenbin Hu, Shaojin Chen et al. (+6 more)
Prefix reuse improves LLM inference by sharing the KV cache of common
prompt prefixes. This helps only if the scheduler can execute a matched branch in the current worker round. Schedulers assume that a matched prefix is ready, which holds for full
attention but fails for hybrid LLMs. There, a branch (the consumer) may need recurrent state that another branch (the producer) is still computing in the same round. The native scheduler stops its waiting-queue scan at this blocked consumer, and the resulting head-of-line blocking creates execution bubbles: idle batch slots that independent work could fill. We introduce RadixFlow, which decouples prefix matching from state readiness. Its state-ready traversal turns a global scan barrier into a local wait: only the consumer is blocked until the next round, while feasible independent branches keep filling the batch under the native state check. On five hybrid models, RadixFlow lowers mean latency by 3% to 24% and worker rounds by 11% to 44% at almost unchanged reuse. On the larger Qwen3.5 models, ablations show that most of the gain comes from traversal, which adds little on a full-
attention control. Heavier rounds limit how much of this saving reaches latency.
Read the full abstract ▾
All authors Huihao Jing, Wenbin Hu, Shaojin Chen, Haochen Shi, Haoyu Huang, Sirui Zhang, Yuxuan Liu, Zhongwei Xie, Haoran Li
Computer Vision
FlowCred: Hierarchical Credit Assignment and Functional Coordination for Joint Text-to-Video Post-Training
Mingxuan Zhu, Yanwei Huang, Shiyuan Liu et al. (+4 more)
Prompt optimization and generator post-training are two complementary ways to improve text-to-
video (T2V) generation, and recent systems increasingly train both modules jointly. However, optimizing a
prompt policy and a
video generator from the same terminal reward creates two ambiguities:
prompt quality is confounded with
stochastic generation, and the two modules may realize overlapping local improvements. We introduce FlowCred, a two-level framework for joint T2V post-training. First, hierarchical credit assignment uses matched
prompt-seed rollout grids to separate
prompt-level credit from
prompt-conditioned generator residuals under shared randomness. Second, functional coordination probes alternative current-policy
prompts at states from actual generation trajectories, providing generator-aware evidence for
prompt optimization while selectively attenuating generator updates already supported by conditioning changes. Across multiple T2V
benchmarks, FlowCred consistently improves matched joint-training baselines. On LTX-2.5, FlowCred improves a strictly matched joint-training control from 61.47 to 64.46 on T2V-CompBench and from 55.58 to 58.08 on StoryEval, without increasing decoded-
video or reward-model-call budgets. Factorial and realized-update analyses show complementary contributions from
prompt shaping and generator routing, while human evaluation and transfer across backbones and updater families provide further supporting evidence.
Read the full abstract ▾
All authors Mingxuan Zhu, Yanwei Huang, Shiyuan Liu, Xinzhi Zhang, Shujie Li, Yuanxing Zhang, Lu Chen
Artificial Intelligence
Mitigating Cross-Modal Hallucinations by Correcting Modality Misrouting in Omnimodal LLMs
Yu Mi, Licheng Pan, Jiayi Zhu et al. (+7 more)
Omnimodal large
language models (OLLMs) are susceptible to cross-modal
hallucinations under audio-visual interference. We find that, for a substantial fraction of incorrect joint predictions, the question-relevant unimodal branch can still recover the correct answer, suggesting that task-relevant evidence can remain recoverable from the relevant modality even when it is not reflected in the joint prediction. We refer to this recoverable failure pattern as modality misrouting. We propose CaRe (Conflict-aware Residual Intervention), a training-free inference-time method for identifying and correcting modality misrouting. CaRe uses a text-only modality probe to identify a candidate modality-specific counterfactual branch, and flags potential misrouting when its prediction conflicts with the original joint prediction. For such cases, CaRe computes a residual direction from the discrepancy between the counterfactual and fused representations and uses it to steer the original
multimodal state toward the relevant modality evidence, while retaining both modalities in the original audio-visual inference path. The intervention strength is adapted based on the confidence of the counterfactual branch. Experiments on CMM and AVHBench across two OLLMs show that CaRe consistently outperforms the evaluated inference-time
hallucination-mitigation baselines.
Read the full abstract ▾
All authors Yu Mi, Licheng Pan, Jiayi Zhu, Qirui Liu, Yitong Qiao, Changtong Xu, Tao Wei, Fei Sun, Zhixuan Chu, Kui Ren
Computation and Language
SparPR: Runtime KV-Cache Recovery for Resource-Constrained LLM Inference on Consumer Hardware
Jieyun Deng, Keming Hu, Yanghe Liu et al. (+1 more)
Running large
language models (LLMs) on consumer hardware is bottlenecked by the key-value (KV) cache, which grows with every decoded
token and quickly exhausts the
memory and bandwidth of a single consumer GPU. We present SparPR (Sparse Partial Recomputation), a runtime KV-cache recovery mechanism that does not modify model weights and requires no offline training. SparPR couples three online decisions: Query-Aware key-block selection scores historical blocks against the current query; dynamic budget control estimates per-block copy and recompute costs to decide how many blocks to retain; and on-demand local recomputation with cost-aware fallback merges retained blocks into continuous spans and switches between sparse recomputation and full recovery according to a runtime cost comparison. Implemented on FlexGen/FlexLLMGen, SparPR is evaluated on a single consumer laptop (RTX 3070 Ti 8GB, Ryzen 9 6900HX, 16GB RAM). It cuts end-to-end latency by 31%, 52%, and 63% at 1k, 2k, and 4k
prompt tokens respectively, with 1.4x-2.7x higher throughput, and completes an 8k-
token workload in 44 minutes where the FlexGen baseline times out. Under a fixed cache budget, SparPR retains a Needle Accuracy that stays well above StreamingLLM and H2O at every keep ratio, and ablation studies confirm that the three mechanisms act jointly.
Read the full abstract ▾
All authors Jieyun Deng, Keming Hu, Yanghe Liu, Li Feng
Machine Learning
Revisiting Error Feedback for Model Quantization
Wai-Yiu Keung, Chi Cui, Wing-Kin Ma et al. (+2 more)
We study causal error-feedback coefficients as design variables for model quantization. Modeling instantaneous quantization errors as independent, zero-mean variables with a common fixed variance, we formulated a metric-dependent trace problem for designing feedback coefficients. The formulation resembles fixed-order GPTQ/LDLQ feedback and supports Jacobian-aware designs that account for downstream error propagation. For a positive-definite metric $\boldsymbol{H}$, we prove that the joint rotation-feedback trace minimum is $n\det(\boldsymbol{H})^{1/n}$ and present an attaining pair using the
geometric mean decomposition. Experiments on synthetic networks, trained MLPs, and
language models show benefits from downstream-sensitive feedback, but also reveal that trace optimality need not improve finite-level performance. Small QAT experiments further illustrate direct loss-driven feedback tuning with fixed or trainable network weights.
Read the full abstract ▾
All authors Wai-Yiu Keung, Chi Cui, Wing-Kin Ma, Jesús Cid-Sueiro, Hei Victor Cheng
Systems and Control
Preliminary Exploration of Large Models Based on Self-Supervised Learning for Health Monitoring of Power Electronic Converters
Xinyuan Liao, Xiangtian Deng, Yang Liu et al. (+3 more)
Current artificial intelligence (AI) methods for power electronics converters make limited use of the abundant unlabeled data generated during real-world operation. This underexplored resource holds significant potential for advancing model performance. To address this gap, this paper presents PE-TSFM, a time-series foundation model tailored for power module health monitoring. The model adopts the PatchTST architecture combined with self-supervised learning and is trained on more than 143 million timestamps, where over 98% of this training data is unlabeled and collected directly from an operating traction converter. In degradation classification tasks, PE-TSFM achieves an accuracy of 87%, considerably surpassing five conventional time-series AI models whose performance ranges from 27% to 45%. These results demonstrate the model's strong few-shot learning capability and generalization across different operating conditions. This work promotes a shift from narrowly focused models in power electronics toward unified, scalable foundation models that can unlock value from large-scale unlabeled operational data.
Read the full abstract ▾
All authors Xinyuan Liao, Xiangtian Deng, Yang Liu, Zhaobing Cai, Haoran Wang, Yi Zhang
Materials Science
Local-resolvent formulation and assessment of linear-scaling Kohn-Sham calculations
Albert Linda
Kohn-Sham density-functional theory is expensive because it needs orbitals that extend over the whole system. For
insulators this can be avoided by using only the diagonal of the one-electron resolvent, $g(\mathbf{r},z)=\langle\mathbf{r}|(z-\hat h)^{-1}|\mathbf{r}\rangle$, at a few complex energies $z$. We call the scheme local-resolvent density-functional theory (LRes-DFT). The density and the band energy are linear in $g$, so neither orbitals nor a kinetic-energy functional are needed. The complex energies are $M$ equally spaced points on a circle that encloses the occupied levels, offset by half a spacing so that no point lies on the real axis. Summing the resolvent over them is equivalent to occupying each level with $1/(1+u^M)$, where $u$ is the level's energy measured from the circle's center in units of its radius. This is a smooth step that is near 1 inside the circle and near 0 outside, and it stays between 0 and 1 for every $M$. If the circle moves with the chemical potential, forces at fixed electron number are exact. Gauss-Legendre and Zolotarev rules need fewer points when the gap is small. In three dimensions we use eigenpairs of small windows around each molecule instead of resolvents. This is divide-and-conquer (DC) with a different occupation function, and it is as accurate as DC with Fermi-Dirac occupations. For H$_2$ molecular
crystals of up to 1024 electrons it agrees with Kohn-Sham to $2.5\times10^{-7}$ Ha per electron. On one core it is slower than dense diagonalization below about 500 electrons and 3.5 times faster at 1024 electrons. It loses accuracy for disordered or compressed systems and for diffuse basis functions.
Read the full abstract ▾
Hardware Architecture
FPGA Based Design And Implementation of Neural Network Compensators for Nanocrystalline MEMS Pressure Sensor
Somsubhra Talapatra, Sudip Ghosh, C Roy Chaudhuri et al. (+2 more)
In this paper the effects of nonlinearity and temperature of nanocrystalline MEMS pressure sensor have been compensated using neural network
algorithm which has been implemented on a FPGA chip. The neural network employs back propagation
algorithm to train the multilayer perceptron feed forward network that causes an error reduction due to temperature upto the level of 98%. It is realized on FPGA with pipelined architecture for parallel and fast sensing operation. The pressure sensor along with the FPGA based compensation results in an intelligent sensor system capable of high speed and accurate pressure measurement which has been tested for a typical biomedical application.
Read the full abstract ▾
All authors Somsubhra Talapatra, Sudip Ghosh, C Roy Chaudhuri, Biplab K Sikdar, Hafizur Rahaman
Robotics
SkipWorld: Latent Skip Planning with Distilled World Models for Efficient Visual Navigation
Yifei Yang, Zehua Fan, Zike Yan et al. (+2 more)
World-model-based visual navigation enables action planning by predicting future states under candidate trajectories. However, long-horizon rollout remains challenging: models are often trained with ground-truth visual histories but deployed in closed-loop prediction, and evaluating many candidates with step-by-step rollout is computationally expensive. We propose SkipWorld, a latent skip planning framework for efficient visual navigation. SkipWorld introduces three components: (i) Latent AutoRoll, which trains a world model to recursively roll out future states in latent space, reducing the mismatch between training and closed-loop inference; (ii) SkipDistill, which distills the latent autoregressive model into a skip world model that directly predicts long-stride future latent states; and (iii) Skip-then-Refine Planning, a dual-model coarse-to-fine planner that first filters candidate trajectories with the skip model and then reranks selected candidates with the stronger autoregressive model. This design preserves accurate trajectory evaluation while reducing redundant rollout computation. Experiments on RECON, HuRoN, SCAND, and TartanDrive show that, compared with NWM, SkipWorld reduces the average ATE by 4.2% and RPE by 2.7%, while decreasing planning time by 62.4% with a 2.7-fold speedup.
Read the full abstract ▾
All authors Yifei Yang, Zehua Fan, Zike Yan, Zhongyu Li, Yan Wang
Solar and Stellar Astrophysics
MATISSE unveils a dusty disc in the interior of EP Aquarii. First evidence of a misaligned hierarchical triple system with a central AGB star?
J. Alonso-Hernández, C. Paladini, C. Sánchez Contreras et al. (+26 more)
Aims. We have mapped the dust and molecular gas emission distributions in the vicinity (a few stellar radii) of the asymptotic giant branch (AGB)
star EP\,Aquarii, with a previously reported stellar companion at $\sim$50\,au, to search for binary-induced asymmetries. Methods. We present an analysis of new high-angular-resolution ($\gtrsim$4\,mas) images obtained with MATISSE (in the wavelength range $\sim$3–8\,\mum) as well as archival high-angular-resolution ($\gtrsim$20\,mas) ALMA $\rm^{12}$CO\,(c$J$=2-1) line and continuum emission observations ($\sim$230\,GHz), to study the circumstellar morphology and kinematics of the molecular gas on these spatial scales. Results. The MATISSE images display a bright central source and two slightly extended emission features ("blobs") located on opposite sides of the AGB
star (at radial distances of $\sim$4–10\,au). These structures are consistent with previous mid-infrared interferometric observations, suggesting stability over decades. The central source is elongated along the direction defined by the blobs, in both the MATISSE and ALMA continuum images. The ALMA $\rm^{12}$CO\,($J$=2-1) maps display an asymmetric inner wind with a velocity gradient. Conclusions. The observations are consistent with an edge-on dusty equatorial overdensity (disc-like), hinting the presence of a close ($\lesssim$4\,au) companion. The inferred
symmetry axis is almost perpendicular to the CO velocity gradient and the orbital plane proposed for the distant ($\sim$50\,au) companion. EP\,Aqr may represent the first misaligned hierarchical triple system with a central AGB
star.
Read the full abstract ▾
All authors J. Alonso-Hernández, C. Paladini, C. Sánchez Contreras, J. Drevon, J. P. Fonfría, L. Planquart, T. Khouri, L. Siess, A. Jorissen, B. Aringer, F. Baron, E. De Beck, S. Höfner, J. Hron, F. Lykou, F. Millour, M. Montargès, N. Nardetto, K. Ohnaka, H. Olofsson, V. Răstău, J. Sanchez-Bermudez, L. Velilla-Prieto, W. Vlemmings, M. Wittkowski, G. Weigelt, W. C. Danchi, F. Kerschbaum, B. Lopez