Search research

We present HERCULES, an open-source simulator and data-collection pipeline for heterogeneous multi-robot autonomy. Built upon the Unreal Engine 5 (UE5)-based simulators AirSim and Cosys-AirSim, HERCULES resolves key architectural limitations of prior frameworks to enable concurrent unmanned aerial and ground vehicle (UAV-UGV) operation in large-scale, photorealistic, dynamic environments. It introduces a new waypoint-tracking UGV controller that mirrors existing UAV control interfaces, and provides a shared navigation stack for mapping, traversability analysis, planning, and control across heterogeneous platforms. Expanding inherited sensor suites, it adds physics-based long-wave infrared (LWIR) cameras and configurable night-vision modes for degraded visual environments. HERCULES provides lightweight APIs, ROS 2 wrappers, and rigorous time synchronization across sensors and platforms, and brings state-of-the-art game-engine capabilities into robotics simulation, integrating intelligent agents such as pedestrians, traffic, and wildlife with high-fidelity dynamic phenomena, including fire, flooding, and crop disease spread. HERCULES runs in two modes: passively, replaying offline-designed trajectories to generate reproducible multi-modal datasets, and actively, running an online planner in closed loop from live observations. Our experiments in heterogeneous multi-robot SLAM, collaborative perception, and exploration, using both HERCULES-generated data and active closed-loop execution, demonstrate its utility for advancing heterogeneous multi-robot autonomy. We publicly release our source code, experiment code, documentation, and datasets, including a heterogeneous multi-robot SLAM benchmark collected with two UAVs and two UGVs across kilometer-scale desert, forest, and city environments, at https://lunarlab-gatech.github.io/HERCULES-website.

Before letting an agent operate over real context, can you prove it used the right evidence? GroundEval turns that question into a deterministic test of what the agent searched, fetched, cited, and was permitted to access. In one case study, two frontier LLM judges scored a plausible agent response 0.85 and higher. But the trace told a different story: the agent had never retrieved the artifact its answer depended on, yielding a GroundEval score of 0.000. We introduce GroundEval, a judge-free framework for evaluating agents against grounded, time-bounded, and access-controlled evidence. GroundEval uses a domain configuration to generate questions, lets the agent choose how to answer, and then scores both the final answer and the recorded trajectory that produced it. The benchmark targets three failures that LLM-as-judge evaluation struggles to detect: whether an agent checked before claiming absence, reasoned only from evidence available to the actor at the relevant time, and used the correct causal mechanism rather than a plausible one. These correspond to three tracks: Silence, Perspective, and Counterfactual. GroundEval exposes when plausible answers rest on invalid evidence paths, and produces structured per-question diagnostics that pair tool activity with the agent's turn-level narration, making each score inspectable rather than merely reported. Our case studies suggest this failure mode is common rather than exceptional, one that final-answer and judge-based evaluation cannot detect by construction.

We present Generative Relightable Avatars (GRA), a person-specific method for photorealistic free-view rendering and environment-map relighting of full-body humans. We postulate that modeling fine-grained appearance details is inherently a one-to-many problem that can benefit from a generative formulation. In contrast to fully regressive relightable avatar methods, GRA follows a hybrid approach that combines controllable, physics-grounded relighting with probabilistic refinement. Starting from a tracked animated mesh, we optimize material parameters in UV-space and render a coarse relit appearance under a target HDR environment map. Next, we refine the textures with a feed-forward model to capture pose-dependent texture dynamics and illumination effects beyond simplified reflectance assumptions. Finally, a fine-tuned video-to-video diffusion model transforms the physically grounded renderings into temporally coherent, high-detail videos while preserving 3D control, with an error-recycling strategy for generating long videos. Experimental evaluations demonstrate our method's improved perceptual quality over prior relightable avatar baselines. Project Page: https://vcai.mpi-inf.mpg.de/projects/GRA/

Self-play with naive gradient ascent cycles in two-player zero-sum games: the last iterate orbits the equilibrium. Modern methods restore last-iterate convergence by regularizing toward a reference policy -- MMD a fixed one (reaching only the regularized equilibrium), R-NaD a periodic snapshot (the engine of DeepNash). We study GARIP, which anchors to the running average, and isolate what the choice of reference controls. Our central result is a mechanism: collapse tracks the peak lag of the reference, and among causal convex averages of a fixed mean lag the running average (flat profile, peak $=$ mean) uniquely minimizes that peak, while a snapshot's sawtooth has peak $= 2\times$ mean (a one-line theorem). Two consequences follow. Convergence: we prove local last-iterate convergence at constant anchor strength -- the anchor scales the base map's rotation by $1-β$, crossing the stability boundary and turning a recurrent base into a contraction (global convergence is conjectured at small $β$; we characterize a large-$β$ consensus failure). Robustness: GARIP matches R-NaD's peak performance -- on matrix games, the Coin Game, and the board games Connect Four/Othello, both moving references are far more robust than fixed-magnet and magnet-free baselines -- but is the better hyperparameter default; we report it both ways: over the full grid collapse rates are statistically indistinguishable, yet at conventional parameterizations a matched-mean-lag setting collapses in 0/40 vs 10/40 seeds (a snapshot matches it only by knowing to shorten $K$). The boundaries: an anticipatory (negative-weight) reference does better still on the stale side, and the advantage appears only where naive self-play cycles (five deep self-play loops). All experiments are pure JAX and reproducible.

Physical neural networks promise low-power machine learning by computing directly with analogue device physics, but most architectures force nonlinear device responses to act as scalar weights. Inspired by Kolmogorov-Arnold networks, we place trainable nonlinear functions on the connections, making each physical connection a learnable computational element. Realising these functions as analogue band-pass filters on field-programmable analogue arrays, we find that the benefit is task-dependent and follows from the smoothness of the physical basis: the networks represent smooth, continuously valued targets, including robotic kinematics, continuous control, and photovoltaic maximum-power-point tracking, with far fewer nodes and connections than multilayer perceptrons, but offer no parameter-efficiency advantage on classification-like decision boundaries. Trained networks transfer to hardware across approximately 35,000 connections with quantified fidelity, and a dedicated CMOS implementation is projected to operate at approximately 30 microwatts. A memristive realisation reproduces the same behaviour in simulation, indicating that the advantage comes from placing trainable nonlinearity on connections, rather than from a particular device.

In this paper, we present Kiwano, an open-source toolkit designed to advance research and evaluation for speaker verification. Kiwano provides a lightweight yet extensible framework built on PyTorch, offering standardized recipes, pretrained models, and integration of several widely used speaker verification architectures. The toolkit emphasizes reproducibility, by delivering transparent training pipelines, unified evaluation protocols and ready-to-use baselines across multiple corpora. Beyond conventional training and inference, Kiwano includes tools for benchmarking, experiment tracking and rapid prototyping of new architectures. To foster community adoption, the toolkit is distributed under the Apache 2.0 license, accompanied by comprehensive documentation and reproducible experiments. By lowering entry barriers and standardizing evaluation practices, Kiwano contributes a valuable resource for both academic research and applied development in speaker verification. The toolkit is publicly available at: https://github.com/kiwano-toolkit/kiwano/

Idling Vehicle Detection (IVD) seeks to determine, at the final frame of a video clip, whether any vehicle is idling, meaning the vehicle is stationary with its engine running, using synchronized video from a remote surveillance camera and multichannel audio captured by spatially distributed wireless microphones along the roadside. Prior full-image, clip-level fusion approaches tend to overfit scene background and full-frame context, produce unstable temporal decisions, and lack an explicit spatial prior to align vehicles with microphones, which makes them brittle under domain shift and data inefficient. Instead, we introduce TAVR-IVD, an audio-visual framework guided by multi-object tracking. Our method detects vehicles, links detections into tracklets, and classifies each vehicle by operating on its tracklet. This design raises the effective signal-to-noise ratio, stabilizes temporal decisions through tracklets, enforces an explicit spatial prior to align vehicles with microphones, and adapts across domains with limited calibration annotations while remaining detector agnostic and efficient. To evaluate deployment robustness, we further curate two evaluation extensions, AVIVD-LT and AVIVD-M, covering inter-day and cross-site shifts.

While longitudinal brain PET imaging is the gold standard for quantifying the spatiotemporal accumulation of Beta-amyloid, its widespread clinical utility is constrained by high operational costs and cumulative radiation risks. Recent deep generative models show promise in longitudinal image synthesis; however, they often fail to capture subtle pathological progression due to identity drift and a persistent bias toward trivially replicating baseline signal intensities rather than modeling temporal transition. To this end, we propose Delta-Diffusion, a novel progression-aware framework that redefines longitudinal PET synthesis as a conditional Poisson Diffusion Bridge (PDB) process. Unlike standard diffusion models that start from Gaussian noise, our PDB formulation is mathematically anchored to the subject's baseline PET, effectively transforming the generative task into a conditional distribution transition of the amyloid trajectory. To handle heteroscedastic nature of PET imaging, we introduce a physically-grounded Poisson perturbation within a Diffusion Transformer (DiT). This architecture uses adaptive scale-shift modulation to precisely calibrate the synthesis with the elapsed clinical interval and structural MRI context. A volume-of-interest balanced objective is designed to emphasize sparse, high-risk regions of amyloid accumulation. Validated on two cohorts with 542 subjects, Delta-Diffusion demonstrates superior performance in capturing longitudinal variations in amyloid deposition compared to state-of-the-art methods, offering a robust computational framework for tracking disease progression.

Dynamic 3D Gaussian splatting faces a fundamental tension between motion consistency and visual fidelity. Deformation-based approaches preserve temporal correspondence but suffer from motion over-factorization, oversmoothing high-frequency dynamics. In contrast, 4D-primitive methods capture fine visual details yet incur temporal overparameterization, breaking object identity and leading to severe storage overhead. To resolve this, we introduce Multi4D, a framework for high-fidelity dynamic Gaussian Splatting based on multi-level competitive allocation. Instead of a monolithic representation, we distribute modeling capacity across three structured levels: static structure, persistent dynamic geometry, and transient appearance primitives. Through shared rasterization and residual-driven optimization, these levels dynamically compete to explain photometric error, enabling adaptive specialization without pre-assigned decomposition. This allocation preserves long-term motion consistency while capturing fine dynamic detail, achieving state-of-the-art rendering quality and real-time performance with significantly fewer dynamic primitives. Furthermore, because our representation explicitly tracks compact persistent Gaussians over time, semantic features can be embedded afterward, enabling Multi4D to achieve state-of-the-art 4D segmentation accuracy with an order-of-magnitude speedup. Project page: https://batfacewayne.github.io/Multi4D.io/

Accurate recognition of anatomical structures is essential for safe and effective minimally invasive surgery (MIS), yet it remains underexplored in surgical computer vision due to limited annotated data and methods tailored primarily to natural scenes. In this work, we present a combined dataset and model framework to advance anatomy-aware perception in MIS. First, we introduce ATLAS-120k, a large-scale clip-level semantic segmentation dataset comprising over 120,000 annotated frames from 100 surgical videos spanning 14 procedures and multiple modalities, including laparoscopic and robot-assisted surgery. The dataset captures substantial procedural variability and was created using a scalable annotation pipeline that integrates expert manual labeling, automated propagation, iterative refinement, and surgeon verification to ensure high-quality annotations. Second, we propose ATLAS (Anatomy Recognition with Context Learning using Foundation Representations), a video semantic segmentation model specifically designed for surgical anatomy recognition. Unlike conventional approaches that emphasize object tracking, ATLAS leverages foundation-model embeddings together with lightweight temporal reasoning to incorporate contextual cues such as procedure type, surgical phase, and short-term visual memory. This design enables temporally consistent and accurate predictions while maintaining real-time feasibility. Together, the dataset and model establish a practical foundation for robust surgical scene understanding and support the development of clinically applicable guidance systems for minimally invasive surgery. The models, dataset annotations and annotation platform are publicly available at: https://github.com/TimJaspers0801/ATLAS.

Cooperative aerial transport requires controllers that respect nonlinear manifold geometry, operate without centralized coordination, and respect operational safety constraints. To address these demands, we present GPAC, a four-layer hierarchical architecture that enables $N$ quadrotors to transport a cable-suspended payload without a central coordinator or by exchanging cable states or adaptive parameters. The key insight is implicit coordination: each quadrotor independently estimates its effective load share from local cable measurements, so combined forces converge to the correct total, even without knowledge of $N$ or the payload mass; the payload position is reconstructed locally from each agent's own cable geometry, and the only inter-agent communication is a low-rate neighbor-position broadcast for collision avoidance. GPAC operates directly on the full nonlinear configuration manifold and integrates geometric position and attitude control, anti-swing regulation, an extended-state observer for wind rejection, concurrent learning-based mass estimation without persistent excitation, and a priority-ordered control barrier function (CBF)-inspired safety filter that reduces operational risk, with input-to-state safety (ISSf) margins that hold exactly under single-constraint activation. A compatibility result shows that the filter's force modifications keep the desired attitude within the almost-global stability region of the $\mathrm{SO}(3)$ attitude controller. Finally, high-fidelity simulation with flexible cables, onboard sensor fusion, and wind turbulence -- with all control and estimation loops closed through the estimator -- yields a mean payload-tracking RMSE of 33.8 cm (2.8% coefficient of variation over 13 seeds) at a low per-agent computational cost.

Tailless flapping-wing micro-aerial vehicles (FWMAVs) mimic the impressive flight performance of hummingbirds, utilising unsteady aerodynamic effects. However, existing designs are still limited and purpose-built with a restricted flight envelope and poor endurance. We therefore propose an adaptable tiltwing framework enabling bioinspired aerial robots to switch between hovering flight, high-speed directional flight, and energy-efficient gliding flight. The proposed framework utilises thrust vectoring with a wide actuation range via two fully independent propulsion units, each flapping a single wing, for effective control and enhanced manoeuvrability. For this, we developed a hybrid Scotch-yoke-based flapping mechanism that ensures a symmetric motion profile with a modular design guaranteeing an arbitrarily wide flapping angle to exploit the lift-enhancing clap-and-fling effect. Additionally, we implemented a passive wing-rotation mechanism, which, in combination with our dual-wing thrust-vectoring approach, allows unprecedented wing-design freedom, unlocking potential for precise optimisation. A contactless leading-edge tracking sensor provides accurate feedback on the wing's orientation and, in the gliding mode, enables dihedral-angle control, augmenting the active wing-pitch control. Extensive testing of a propulsion unit was conducted with a six-axis force/torque sensor, demonstrating the flapping mechanism's performance while optimising transmission efficiency and the passive wing-pitch mechanism. At full throttle, the average lift force generated by a single wing, flapping with a 188° amplitude, was 21.1 gf for a small 3.1 g 1S BLDC motor. Additional tests covering the full range of the wide-angle tilting capability showed an effective thrust-vectoring control architecture with a linear and symmetric response curve of the moments generated.

Fingerprints are the most widely deployed biometric. Verifying whether two impressions come from the same finger typically relies on minutiae, small landmarks such as skin ridge endings and bifurcations. These landmarks are extracted through a multi-stage pipeline of image enhancement, skeletonization, minutiae detection, and alignment. We investigate an alternative: using topological data analysis to represent the full pattern of skin ridges and valleys directly, bypassing minutiae detection and the downstream matching pipeline. We apply persistent homology, a topological tool that tracks how loops in the ridge pattern form and fill in across spatial scales, producing multi-scale summaries of ridge geometry. We develop and compare a range of verification methods on a standard benchmark dataset, FVC2000 DB1. Even the simplest topological summaries, with no trained parameters, substantially outperform geometry-only baselines. A trained method achieves an AUC of 0.91, while an optimal-transport method excels at the strictest false-accept thresholds, suggesting they capture different aspects of the ridge pattern. Fusing these two approaches yields the best performance at every low false-accept threshold we examine. Our results establish that these topological summaries capture substantial fingerprint identity information, far more effective for verification than raw pixel-level geometry. Because the entire pipeline is openly specified, it offers a transparent complement to minutiae-based systems, and we provide a modular framework for constructing, evaluating, and combining topological verification methods.

Online planning in continuous partially observable Markov decision processes (POMDPs) using $ω$-regular specifications requires handling continuous belief dynamics within the finite symbolic memory in order to track temporal progress. Existing methods based on either direct search in belief space or predefined discrete abstractions suffer from drawbacks, e.g., lack of symbolic memory for long-horizon logical progress or difficult to certify from noisy online beliefs. As such, obtaining reliable symbolic states online from continuous observations remains a challenge. To address this issue, we introduce the Revealed Belief Automaton (REBA), an event-driven framework that advances the research from global belief-space discretization to a fundamental new way of thinking, namely online certification of revelation events. Specifically, we propose an online revelation method that, through information-theoretic gates, can dynamically analyse and establish belief abstraction from the continuous belief space by discovering reliable anchors among noisy beliefs. We then develop an incremental topology adaptation mechanism over the certified anchors to realise the online finite Belief Automaton. By combining with the $ω$-regular specification, REBA is able to support formal parity policy synthesis without a predefined discrete abstraction, which in turn can guide the Monte Carlo Tree Search process to perform online search beyond its local horizon. In addition, we design an error decomposition analysis which can assess the effectiveness and reliability of this discrete guidance for the underlying continuous POMDP. Empirical evaluations in patrolling and navigation scenarios show that REBA matches or exceeds all evaluated baselines, with primary metric gains of +17.0% to +47.4% over state-of-the-art approaches.

Recent advances in text-to-speech (TTS) have greatly improved speech naturalness, speaker similarity, and controllability. However, most existing controllable TTS systems still rely on explicit user-provided style prompts, making it difficult to automatically determine how a sentence should be spoken in long and complex conversational scenarios. This proposal introduces the ISCSLP 2026 CoT-TTS Challenge, which aims to evaluate whether a system can infer the intended speaking manner from contextual information and generate speech consistent with both the reasoning output and the surrounding scene. The challenge contains two tracks: text-context-aware CoT-TTS and audio-context-aware CoT-TTS. We construct a large-scale bilingual training set from speech-rich media and provide carefully filtered evaluation data for leaderboard comparison. Each system is required to output both a chain-of-thought reasoning analysis and the generated speech waveform. The official evaluation combines objective metrics, multimodal LLM-based evaluation, and human subjective assessment. To facilitate reproducibility, we provide inference code together with a fine-tuning recipe for a 0.6B Qwen3-based model trained via a three-stage strategy. This challenge is expected to support research on context understanding, chain-of-thought reasoning, and expressive speech generation for applications such as film dubbing, audiobook production, virtual characters, and spoken dialogue agents. Further information about the associated challenge is available at:https://iscslp2026-cot-tts.github.io/challenge-website/

The Categorical Jacobian of Zhang et al. (2024) reads protein contacts from a language model by perturbing every residue with every alternative amino acid, about $19L$ forward passes. We show the signal it reconstructs is already concentrated in a small subset of attention heads: averaging the top-$K$ contact-relevant heads -- selected on as few as 10 labeled proteins, with no fitted per-pair or per-head weights -- recovers contacts in a single forward pass and matches or beats the Categorical Jacobian for every bidirectional model where it is defined (bar the smallest, 8M). Our primary test is leakage-clean: on a CAMEO split where neither selection nor evaluation touches data the models have plausibly memorized, the head readout beats the Categorical Jacobian on ESM-2-650M by +9pp ($N = 29$, $p < 0.001$), with the within-model margin reproducing across architectures. Ablations localize the gain to labeled head selection, not to averaging: at a matched label budget the unweighted mean ties a supervised $L_1$ logistic regression on the same heads. Both methods fall 30-36pp from their in-distribution Zhang numbers to the leakage-clean split, which we read as an upper bound on how much prior numbers reflect pretraining overlap. We additionally introduce representation-CJ, a hidden-state generalization of the Jacobian to architectures without a masked-LM head (the output-head-independent analogue of logit-CJ), agreeing with the Categorical Jacobian where both are defined (per-protein Pearson $r \approx 0.95$); show that the optimal $K$ tracks how diffusely a model spreads its contact heads; and find both methods lose the signal on the two causal LMs we test, suggesting attention-encoded pair structure may depend on bidirectional pretraining.

Fish achieve efficient swimming across varied speeds through active modulation of their body flexibility. To explore the effects of tunable stiffness on swimming performance, we present a bio-inspired freely swimming fish robot with a rapidly tunable particle-jamming body. This design enables rapid stiffness adjustments with negligible changes in shape or volume, achieving a 54% variation in flexural rigidity across vacuum pressures of 0 to -40 kPa. We visualize the midline of the oscillating body under both low- and high-stiffness conditions, and the comparison confirms that the body curvature varies with stiffness. We further experimentally evaluate the tunable stiffness body's effects on swimming performance using velocity and cost of transport (CoT) measurements obtained via a motion tracking system. Results show that active stiffness tuning is essential for sustaining efficient and high-speed swimming across beating frequencies of 1-3 Hz. At low frequencies (1-1.5 Hz), a softer body (0 kPa) maximizes velocity and minimizes CoT, whereas at high frequencies (2.5-3 Hz), a stiffer body (-40 kPa) delivers superior velocity and reduced transport cost. These findings highlight stiffness modulation as a key strategy for adaptive and efficient propulsion in bio-inspired robotic swimmers.

The Morse index of a free boundary minimal surface is encoded in its Jacobi-Steklov spectrum, and we test how faithfully a physics-informed neural network (PINN) reproduces that spectrum on a problem whose answer is already known in closed form. The benchmark is the critical catenoid in the unit ball $\mathbb{B}^3$, where it is well known that the Morse index equals $4$ and the nullity equals $2$. Separating the angular variable reduces the eigenvalue problem to a family of one-dimensional Robin problems on $[-T,T]$, one for each Fourier mode. A network that enforces the parity of each mode by construction, and carries the eigenvalue as a trainable parameter, returns the three eigenvalues below the stability threshold to within $10^{-6}$ to $10^{-4}$ of their exact values, with PDE residuals of order $10^{-4}$; assembling them recovers the index $4$ and the nullity $2$. We then track the spectrum along a one-parameter homotopy joining a flat reference operator to the catenoid Jacobi operator and identify the crossings at which the index changes. Since the critical catenoid is rigid, a fact we prove, this homotopy deforms operators rather than surfaces. We close by explaining how the same pipeline, with its one-dimensional solver replaced by a two-dimensional one, is poised to address genuinely geometric families in ellipsoidal balls, where the boundary curvature is no longer constant, and the Morse index is not yet known.

We describe our entry to the efficiency track of the Academic Text-to-Music (ATTM) Grand Challenge at ICME 2026. Beyond the challenge protocol's FAD-CLAP and CLAP score, we add a learned human-preference reward from TuneJury, a twin pairwise ranker trained over open music-preference datasets. The reward serves both as a training-time conditioning signal and as a sample-selection criterion. The pipeline combines five engineering decisions on a 120M-parameter FluxAudio-S backbone, four at training time and one at inference: (i) training-time reward conditioning that doubles as an inference-time CFG axis, (ii) a sweep over five score-conditioning architectures, where training and inference use different variants, (iii) expert iteration on the top decile, (iv) a short preference-tuning pass (CRPO) for audio-text alignment, and (v) inference post-processing via joint CFG, source separation, and loudness normalization. Per-stage decomposition on 100 Song Describer prompts shows training-time reward conditioning as a functional conditioning axis, expert iteration as the dominant contributor, the preference-tuning pass adding only noise-level gain, and the inference-time score scalar already saturated by the end of the chain.

Robust locomotion and navigation for legged robots relies heavily on dependable odometry. Traditional multi-sensor fusion for such state estimation requires meticulous sensor calibration and platform-specific kinematic modeling, which complicates deployment. Industrially packaged exteroceptive sensors can provide accurate motion tracking but remain vulnerable to perceptually degraded conditions. We thus develop a plug-and-play, robust legged odometry system that eliminates the need for explicit exteroceptive-to-proprioceptive calibration or system kinematic modeling. Our approach leverages established exteroceptive motion pipelines as a continuous supervisory signal to train an online learned velocity neural network directly from proprioceptive data. An Invariant EKF (InEKF) is then used to fuse the learned proprioceptive or exteroceptive velocity (if any) and IMU data. When exteroception fails due to environmental degradation, the system seamlessly falls back to using the learned proprioceptive model, yielding a resilient legged odometry that readily adapts to new hardware. We demonstrate the platform-agnostic, easily deployable nature of our approach on different quadruped platforms, showcasing promising results in maintaining robust motion estimation across challenging scenarios.

Existing embedding models are inherently static: they encode text segments in isolation, ignoring their surrounding context and temporal order. This paper introduces EvoEmbedding, a novel embedding model that generates evolvable representations for retrieval. It is tailored for long-context scenarios, where information is dynamic, sequential, and requires continuous state tracking. Our design is simple: EvoEmbedding maintains a continuously updated latent memory as it sequentially processes inputs, and uses it alongside the raw content to jointly generate evolvable embeddings. Consequently, for the same query, our model adapts its representation to retrieve distinct targets based on the evolving context, going beyond static semantic search. To equip the model with this capability, we construct EvoTrain-180K, a diverse dataset for the joint optimization of latent memory and retrieval. Furthermore, we introduce a memory queue to prevent representation collapse during recurrent encoding, alongside segment-batching techniques that tackle significant length variance and accelerate training by 3.8$\times$. Extensive experiments show that our model not only outperforms larger-scale specialists (e.g., Qwen3-Embedding-8B and KaLM-Embedding-Gemma3-12B) across a range of long-context retrieval benchmarks, but also generalizes well to downstream tasks (e.g., personalization) with contexts 10$\times$ longer than its training window. Notably, EvoEmbedding seamlessly integrates into agentic workflows to boost performance. For instance, a naive RAG pipeline equipped with our model surpasses dedicated agentic memory systems. Project Page: https://clare-nie.github.io/EvoEmbedding/.

Verification of multi-agent systems requires the ability to check meticulous topological properties when it comes to agents that can move through space in continuous time. This demands a logic with sufficient expressiveness to capture these dynamics. MuTGL logic has interesting properties for expressing entangled space-time properties. However, this logic lacks the expressivity needed to analyse reachability within specific distance bounds, or to track the length or the cost of communication chains: these are fundamental for decentralized monitoring, or graph-theoretic analysis of distributed protocols, where algorithmic complexities often relates with the system's communication graph diameter. We then introduce an extension of muTGL, including a new operator called the space horizon. This addition allows us to bound the distance of communication chains, hence enhancing the logic's expressiveness. We show that this operator allows to encode modalities from other logics, such as reachability or escaping which were not available in vanilla muTGL, while allowing a deeper entanglement of spatial and temporal properties. We provide a centralized offline monitoring algorithm for this logic and illustrate it on several examples on simulations of Consensus-Based Bundle Algorithms, distributed protocols for task allocation.

Collaborative Problem Solving (CPS) is a core skill in education, where the process of peer interaction is highly important. However, existing educational dialogue datasets mostly focus on classroom instruction or tutoring (i.e., teacher/tutor-student interaction), yet datasets centering small-group, student-student interaction are limited. This thus leaves research with limited resources for studying how students interact, coordinate, and solve problems together in real educational settings. To address this, we introduce PeerMathDial, the first dataset of peer CPS dialogues collected from authentic middle school math classrooms. It contains 55 dialogues from 27 students, totaling 6,406 turns. To facilitate research on CPS discourse analysis, we further build a corpus-grounded dialogue act taxonomy assisted by LLMs. Using the dataset and the dialogue act taxonomy, we demonstrate the practical applications of PeerMathDial across three use cases. First, we track how dialogues evolve over time and measure the impact of teacher interventions. Second, we align dialogue actions with student surveys to reveal the connection between students' traits (e.g., confidence, leadership) and their actual behaviors. Third, by evaluating LLMs on dialogue act prediction, we glimpse at the potential of LLMs for student simulation in educational applications. Our dataset and source code will be released to the community.

Mixture-of-Experts (MoE) language models are often described as ideal for resource-constrained inference. Each token activates only a small subset of experts, so the per-token compute cost, in floating-point operations (FLOPs), resembles that of a much smaller dense model. Whether that FLOP advantage survives in practice is far less clear. We ask whether MoE models actually run faster and cheaper than comparable dense models on consumer-grade and edge hardware. We benchmark OLMoE-1B-7B (1.3 B active of 6.9 B total) against three dense baselines on an Apple M2 Pro and an NVIDIA Jetson Orin Nano 8 GB through llama$.$cpp, measuring throughput, memory, and on-device energy. The answer is device-dependent: OLMoE's active-parameter advantage is only partly realised on the laptop (~10% behind the same-active Llama-3.2-1B) and erodes on the edge device (~31% behind, at 2.1$\times$ the energy per token, with peak memory at the 8 GB ceiling). Patching llama$.$cpp to time the decode graph node-by-node shows routing accounts for under 9% of MoE-block compute on the cleaner edge backend, so the gap reflects total-parameter memory footprint, expert dispatch, and KV-cache pressure rather than routing. The implication is that on bandwidth-bound edge hardware, inference cost tracks total parameters, not active ones, and sparse activation does not buy back what the device is constrained on. These findings are bounded to one MoE model at this parameter scale and two devices, and we release the full measurement harness and per-run data.

Trajectory prediction allows autonomous vehicles to anticipate the future behavior of surrounding objects (or agents) and, accordingly, maximize the safety and efficiency of their driving. State-of-the-art Transformed-based interaction-aware trajectory prediction models, which rely on attention mechanisms to capture multi-agent interactions and maximize prediction accuracy, are commonly trained and evaluated on long-range high-quality datasets. These datasets are typically obtained by aggregating data from multiple vehicles or drones and removing any object detection or tracking noise offline. Yet, information about a surrounding object's state (its position, speed, heading) is far from being noiseless in real-world deployments. Object state estimation is affected by perception uncertainties and localization errors that can be particularly large for objects received via Vehicle-to-Everything (V2X) communications. In this paper, we analyze the impact of noisy object state information on the trajectory prediction accuracy of a state-of-the-art Transformer-based interaction-aware trajectory prediction model. Our study demonstrates that trajectory prediction accuracy can rapidly deteriorate as the noise intensity increases. Numerical results show that the prediction accuracy can reduce by a 1.3x factor under small noise levels and by as much as a 3.9x factor under the highest (yet realistic) noise conditions. These findings reveal the strong sensitivity of trajectory prediction models to noisy data, underscoring the need for more realistic training and evaluation datasets as well as noise mitigation strategies.

Reliable localization is essential for intelligent transportation systems (ITS), including autonomous vehicles, quadruped last-mile carriers, and infrastructure-inspection unmanned aerial vehicles (UAVs). Although tightly-coupled multi-sensor fusion improves accuracy in favorable conditions, deployed systems remain vulnerable to sensor degradation -- poor illumination, LiDAR degeneracy, wheel slippage, and GNSS outage -- and to spatiotemporal calibration errors. These failures are common in urban canyons, tunnels, and high-speed corridors, where localization drift can degrade route tracking, tunnel passage continuity, and local map alignment. This paper presents Ultra-Fusion, a tightly-coupled multi-sensor localization framework based on a unified sliding-window estimator. Asynchronous measurements are timestamp-ordered and converted into optional factors within one optimization window, supporting WIO, VIO, LIO, and LVIO with optional wheel and GNSS augmentation. Observability-aware initialization selects the bootstrap mode, factor-wise reliability scheduling gates degraded measurements, and online LiDAR--IMU spatiotemporal calibration refines temporal offsets and rotational extrinsics during operation. We extend the M3DGR benchmark with simulation trajectories and evaluate more than 60 open-source SLAM systems on M3DGR, M2DGR-Plus, KAIST, GrandTour, and MARS-LVIG. The results show competitive accuracy across wheeled, legged, and aerial platforms under long-duration and high-speed operation, degradation, and calibration perturbation, improving localization availability for road-level autonomy, campus and warehouse mobility, and low-altitude aerial inspection. To benefit the industrial and academic community, we will release source code and datasets upon paper acceptance.

Long horizon, contact-rich manipulation is inherently partially observable. This is as a single visual observation rarely captures a robot's full action context, including prior attempts, interactions, or progress. Consequently, standard visuomotor policies or vision-language-action models are prone to struggle in such tasks due to a lack of memory. To address this, we introduce Compressed Action Memory Policy (CAMP) based on the insight that a robot's own action history serves as a highly informative, self-supervised signal, enabling the policy to learn a robust, compact history representation. In our approach, we train a memory module to maintain a compressed representation of past actions, forcing it to encode a latent behavioral memory of all the robot's past interactions that can then be used to better contextualize future actions. This allows our approach to implicitly track generalized task progress and learn from failed attempts without any additional supervision, or external oversight. We evaluate CAMP across four real-robot setups and two novel simulation benchmarks: Memory-T-Bench and Memory-Manip-Bench. By demonstrating substantial gains over state-of-the-art baselines, CAMP is, to our knowledge, the first policy to demonstrate substantial success on contact-rich partially observable manipulation tasks purely through learned memory.

3D human avatars have shown impressive visual fidelity driven by pose-conditioned models, yet they still lack the physical ability required for interactions with each other and environments. Although recent studies have made various attempts to incorporate physical characteristics into 3D avatars, they only exhibit limited physical deformations, often leading to constrained interaction behaviors. To resolve this issue, we present PIAvatar, a framework to simultaneously enable physically aware interactions between avatar-avatar and avatar-environment, and a non-rigid deformable human body simulation. In this work, our key insight is to decouple kinematic velocity from deformation gradient. When external forces act on avatars, the kinematic velocity induces stress which hinders the avatar's ability to achieve a desired pose. In addition, we integrate a skeletal framework within the avatar. It allows estimating its poses and real-time tracking in a closed form, even during non-rigid physical interactions. Our approach is implemented within a conventional Material Point Method framework to ensure physically consistent dynamics. We lastly evaluate the method on both human-object and human-human interaction scenarios to assess its behavior under diverse interaction settings.

Decentralized bilevel optimization (DBO) provides a powerful framework for multi-agent systems to solve local bilevel tasks in a decentralized fashion without the need for a central server. However, most existing DBO methods rely on lower-level strong convexity (LLSC) to guarantee unique solutions and a well-defined hypergradient for stationarity measure, hindering their applicability in many practical scenarios not satisfying LLSC. To overcome this limitation, we introduce a new single-loop DBO algorithm called diminishing quadratically-regularized bilevel decentralized optimization (DUET), which eliminates the need for LLSC by introducing a diminishing quadratic regularization to the lower-level (LL) objective. We show that DUET achieves an iteration complexity of $O(1/T^{1-5p-\frac{11}{4}τ})$ for approximate KKT-stationary point convergence under relaxed assumptions, where $p$ and $τ$ are control parameters for LL learning rate and averaging, respectively. In addition, our DUET algorithm incorporates gradient tracking to address data heterogeneity, a key challenge in DBO settings. To the best of our knowledge, this is the first work to tackle DBO without LLSC under decentralized settings with data heterogeneity. Numerical experiments validate the theoretical findings and demonstrate the practical effectiveness of our proposed algorithms.

Large language models are increasingly proposed for mental-health applications such as detecting suicidal content, raising the question of what they rely on. We study this mechanistically and use it to ask a narrower question: how to make a causal claim about a model's internal features more trustworthy. Our validation-gated framework, with suicidality detection as a case study, interprets a behavior only after the model is shown to perform it: a concept is admitted only once the model ranks it above a simple lexical baseline, and each subsequent property is tested against a matched control. This discipline yields negative as well as positive results. The gate rules out one task at the outset: on DeepSuiMind (Li et al. 2025), Llama-3.1-8B-Instruct cannot separate implicit suicidal intent from ordinary distress, so we do not analyze it. We turn to binary suicide detection, which it does perform. There we find a mid-network feature that appears semantic rather than keyword-based, is causally implicated in the decision (ablating it degrades the judgment; a random direction does not), is low-rank, and recurs across three model families and three suicide datasets. A register-matched control (suicide versus depression) suggests it tracks suicidality more specifically than general distress. Steering raises the model's response, but for unrelated questions too, so we treat it as necessary but not sufficient. The clearest pattern separates encoding from use: smaller models already represent suicidality, yet only larger ones appear to act on it. The positive evidence is English Reddit text, which limits the clinical reading.