arXiv:2607.16076v1 Announce Type: new
Abstract: Multimodal sarcasm and cyberbullying detection remain challenging because the intended meaning often emerges from incongruity between textual and visual information rather than from either modality alone. Existing multimodal approaches primarily rely on feature fusion or cross-modal attention, which may not effectively capture hierarchical semantic inconsistencies across different levels of representation. To address this limitation, this paper proposes HCIG (Hierarchical Cross-modal Incongruity Graph Network), a novel framework that models cross-modal incongruity at token, phrase, and global levels using graph attention networks and adaptively integrates these representations through a learned hierarchical attention mechanism. As a complementary architecture, we also introduce GCCN (Graph-based Cross-modal Contradiction Network), which performs graph-based reasoning using contradiction-aware pooling for efficient multimodal interaction learning. The proposed models are evaluated on the MMSD sarcasm benchmark and the MultiBully cyberbullying dataset, together with comprehensive ablation studies and cross-task transfer experiments. Experimental results demonstrate that HCIG achieves the best performance on MMSD with 85.74% accuracy and 85.29% macro-F1, while GCCN attains the highest macro-F1 (68.66%) on MultiBully and HCIG achieves the highest accuracy (69.62%) and bullying-class F1 (74.90%). The findings demonstrate that hierarchical multi-granularity incongruity modeling provides more effective multimodal reasoning than conventional fusion strategies, offering a robust framework for sarcasm and cyberbullying detection in social media.
Science Journals
arXiv:2606.18824v2 Announce Type: replace
Abstract: Pedestrian trajectory prediction from an on-board ego-centric camera is challenging since it depends on complex interactions with vehicles and scene context, as well as the intention of the pedestrian. The task becomes even more challenging since pedestrian intention is often ambiguous from historical observations alone, leading to an inherently multimodal distribution over future trajectories. Existing CVAE-based predictors learn a latent representation by aligning the prior conditioned on trajectory history with the posterior conditioned on both history and future trajectories. This alignment may suppress multimodality in the latent space, leading to mixed-mode predictions that interpolate between distinct future behaviours. In this paper, we propose MMPM, a mode-aware framework that separately models future trajectory distributions into semantically meaningful modes based on the pedestrian's crossing behavior. MMPM consists of two modules: behavior-aware Pedestrian Interaction Module (PIM) that jointly captures pedestrian-vehicle and pedestrian-environment interactions by introducing gaze and hand gestures, and a CVAE-based Mode-aware Trajectory Predictor (MTP) module to model the future trajectory distributions in two modes, crossing and non-crossing the road, separately. A query-based decoder further enforces mode consistency during decoding. Experiments on PIE and JAAD datasets show that our method surpasses state-of-the-art baselines. Our proposed MTP is model-agnostic, which can be integrated into existing frameworks such as BiTrap and SGNet to further improve future trajectory prediction performance. We additionally introduce a data-driven validation protocol that retrieves spatio-temporally consistent real-world trajectories from the same scene and evaluates predictions against these retrieved examples, demonstrating improved frame-wise displacement errors over previous work.
arXiv:2606.24275v2 Announce Type: replace
Abstract: We compare two geometric stochastic frameworks for the two-dimensional Euler equations, being the circulation-preserving stochastic advection by Lie transport (SALT) and the energy-preserving stochastic forcing by Lie transport (SFLT) approaches. While preserving both circulation and energy is ideal, their simultaneous conservation restricts perturbations to a stochastic reparametrization of time. Consequently, a fundamental choice must be made between preserving structure or the kinetic energy. Analysis reveals that SALT is significantly more sensitive to high-frequency flow components, with noise effects scaling by $| \bk |^2$ relative to SFLT. This suggests that SALT acts as a localized perturbation sensitive to sharp gradients, while SFLT behaves as a more regularized global forcing. Numerical experiments on a traveling dipole, vortex merger, and forced-damped turbulence confirm that SALT introduces uncertainty localized near dynamically active vorticity gradients, whereas SFLT produces a more diffuse variance field spread across the domain. These results illustrate how the choice of geometric invariant fundamentally determines scale-sensitivity and spatial distribution of modeled uncertainty in vortex dynamics.
arXiv:2606.24769v2 Announce Type: replace
Abstract: Even when Dirac-Frenkel dynamics determine a well-defined evolution in function space, the corresponding parameter dynamics can be non-unique or ill-conditioned for redundant nonlinear parametrizations, such as typical neural networks or mixture models. We propose to add inertia to the Dirac-Frenkel dynamics and show that this allows useful parameter velocity information to persist from the past trajectory in directions that are weakly informed, while well-informed parameter velocity directions continue to follow the Dirac-Frenkel dynamics. We prove that the inertial formulation yields well-posed parameter dynamics and provide a posteriori error bounds. After time discretization, the method requires the solution of the same type of regularized linear least-squares problem as standard Dirac-Frenkel dynamics, but with the previous velocity appearing as an anchor. Numerical experiments demonstrate the increased robustness obtained with inertia.
arXiv:2606.26743v2 Announce Type: replace
Abstract: Point clouds are an important carrier of three-dimensional spatial information, and their quality directly affects the performance of downstream perception tasks such as object detection and tracking. However, millimeter-wave radar point clouds are typically sparse, noisy, and structurally incomplete. To address these limitations, this paper proposes a multimodal point cloud generation method based on vision-radar fusion. The proposed method leverages image semantic information to impose structural constraints and achieve spatial alignment for radar point clouds, while incorporating a sparse completion strategy to enhance point density and recover missing structures. The generated point clouds are further evaluated in object detection and tracking tasks. Experimental results demonstrate that the proposed method effectively improves point cloud quality and enhances the detection accuracy and robustness of perception models in complex environments, providing a practical solution for multisensor point cloud generation and intelligent perception systems.
arXiv:2606.27400v2 Announce Type: replace
Abstract: We revisit Feynman's classic carousel problem (\textit{Feynman Lectures}, Vol.~1, Sec.~19.4), in which a student walks radially inward on a platform driven at constant angular velocity $\omega_0$ by an external motor. Working consistently in the frame co-rotating with the platform, we show that the student's kinetic energy in that frame is exactly \emph{constant}: the mechanical work done by the real radial friction force is exactly cancelled by the work done against the centrifugal inertial force. The Coriolis inertial force, and the equal-and-opposite tangential friction force required to cancel it and keep the student on a straight radial path, do \emph{zero} mechanical work, because both are perpendicular to the radial velocity. Yet generating that tangential force costs metabolic energy -- precisely the effect Feynman flagged with ``one has to lean over and push sidewise.'' The mechanical model adopted here (a point mass with kinematically imposed trajectory) is deliberately silent about internal physiology; the metabolic cost discussion of Sec.~\ref{sec:cost} is a separate phenomenological layer, explicitly flagged as such, not derived from the mechanics. We give an order-of-magnitude metabolic estimate, an entropy-production argument connecting the exercise to the second law, a feedback (PD-controller) reformulation of the same physics, and a playground experiment students can run with a phone and a heart-rate strap. Throughout, we are explicit about reference frames and about the limits of each model.
arXiv:2606.29123v3 Announce Type: replace
Abstract: In aided inertial navigation, measurements from different sensors are often subject to unknown relative time delays. Consider a single aiding sensor whose measurements have an unknown but constant delay relative to the inertial-measurement data stream. We study the identifiability of the delay and the initial navigation state parameterizing the trajectory. Identifiability depends on both the temporal structure of the aiding measurements and the form of the trajectory. Using the special Galilean group, we determine the minimal number and type of aiding measurements needed to recover the delay and the navigation state. We also characterize a class of \emph{uninformative} trajectories, for which the delayed measurement model admits a continuous symmetry that prevents unique delay-and-state recovery. We show that each such trajectory is generated by a constant element of the Galilean Lie algebra, and connect this result to the familiar linearized, Jacobian-based analysis.
arXiv:2606.29502v2 Announce Type: replace
Abstract: Skill memories can improve agentic reinforcement learning by reusing past experience as textual guidance, but retrieved skills are not oracular: they may help in one state while misleading the same policy in another. This makes the common privileged-teacher assumption fragile, namely that a skill-conditioned prompt can be treated as a fixed teacher for the no-skill prompt. We introduce UCOB, a framework for learning to utilize and evolve agentic skills via credit-aware on-policy bidirectional self-distillation. UCOB treats skill-conditioned and no-skill prompts as two on-policy context views of the same model, compares their return-to-go within the same task and anchor state, and uses the higher-return view as the local teacher. This local credit signal internalizes useful skill-conditioned behavior, corrects misleading skill usage, and guides task/state skill memory updates, utility-aware retrieval, and reflection self-training. Experiments on agentic tasks, including ALFWorld, WebShop, and Search-QA, show that UCOB outperforms skill-free RL, skill-memory baselines, and self-distillation methods across model scales, with up to 23.5 and 18.0 point gains over SOTA baselines on ALFWorld and WebShop. Ablations and analyses further validate its core mechanisms, continual adaptation across environments, and modest training overhead. Code is available at https://github.com/TU2021/UCOB.
arXiv:2606.29519v2 Announce Type: replace
Abstract: Long-range learning is hard for recurrent networks trained with stochastic gradient descent, because the influence of a past input fades with the lag $\ell$, and if it fades too fast the dependence cannot be learned from finite data. This fade is captured by an envelope $f(\ell)$. An exponential fade makes the data needed to learn a lag-$\ell$ dependence grow exponentially, putting long horizons out of reach; a power-law fade keeps the cost polynomial. We show that the asymptotic decay class of $f(\ell)$ is not fixed by the architecture. Instead, it emerges from the coupling between the state dynamics and parameter dynamics, settling into either a collapsed regime (fast, exponential forgetting) or an extended, anti-collapsed regime (slow, power-law forgetting). The intuition is a competition within these coupled dynamics. Training drives the network's effective time scales toward short ones, while rare, heavy-tailed fluctuations of the learning dynamics push a few of them to very long values. The extended regime survives only when these heavy-tailed pushes are strong enough to balance the pull. We make this mathematically precise with a coarse-grained stochastic process and prove exactly when the extended regime exists. A single exponent, the spectral exponent~$\beta$, then governs both the spread of time scales and how slowly the network forgets. Realizing the regime in practice needs one more ingredient: the joint action of the architecture and the optimizer must be able to hold such a broad spread. A network whose capacity to generate broad time-scale spectra is severely constrained still collapses, even when supplied with strong heavy-tailed forcing. Heavy-tailed fluctuations thus act not as noise to be suppressed, but as the mechanism that sustains long-range learning.
arXiv:2606.29538v4 Announce Type: replace
Abstract: Skills are a useful abstraction for software agents, turning human and agent experience into reusable procedural knowledge. Yet existing skill libraries are mostly hand-written, text-centric, or derived from agent traces, leaving tutorial videos and other multimodal human resources largely underused. We present RESOURCE2SKILL, a framework that distills multimodal resources, including tutorial videos, repositories, articles, and reference artifacts, into executable skills for software agents. RESOURCE2SKILL organizes these skills as a hierarchical multimodal Skill Wiki, where each entry combines structured text, code, visual examples, metadata, and provenance. This design preserves complementary signals from different resources: videos capture temporal operations and visual effects, code captures executable tool patterns, and articles or artifacts provide conceptual and stylistic grounding. At inference time, agents retrieve and compose relevant skills from the wiki; when coverage is insufficient, the same construction operator can acquire new skills online. Across seven practical authoring domains, RESOURCE2SKILL improves average overall score by +11.9 percentage points over no-skill agents and outperforms strong harness baselines in 26 of 28 main-aggregate model-domain cells. Ablations confirm the value of multimodal skill format, hierarchical organization, source diversity, selection strategy, and online acquisition.
arXiv:2606.29573v3 Announce Type: replace
Abstract: Multimodal large language models (MLLMs) are increasingly expected to generate fine-grained descriptions of visual content. However, we observe and theoretically show that generating fine-grained responses poses a reliability challenge, \textit{i.e.}, fine-grained generation is more error-prone than coarse-grained generation. This phenomenon suggests that models should generate the finest description that remains reliable rather than simply produce more specific outputs. To investigate this problem, we develop \textsc{GranFact}, a granularity-aware benchmark consisting of expert-verified multi-object images with coarse-to-fine category annotations. Then, we design a hierarchy-aware evaluation algorithm, which assesses both whether model predictions are visually correct and how specific the correct predictions are. We also propose a reliability-prioritized preference optimization method based on Direct Preference Optimization, which penalizes unreliable fine-grained claims while rewarding reliable specificity. Experiments on \textsc{GranFact} show that our method improves fine-grained generation while preserving reliability. Code and data are available \href{https://github.com/WeiWu2025/GranFact}{here}.
arXiv:2606.30072v2 Announce Type: replace
Abstract: Cooperative tasks in Multi-Agent Reinforcement Learning (MARL) require agents to collectively maximize a shared return. Under the Centralized Training with Decentralized Execution (CTDE) paradigm, policy gradients have remained difficult to compute directly. Prior methods largely follow two approaches: independent factorized updates with centralized critics, which lack general joint-improvement guarantees without value decomposition assumptions, or alternating best-response updates, which can converge to suboptimal Nash Equilibria. In this paper, we show the joint policy gradient admits an exact decentralized decomposition of per-agent terms, each formed from per-agent score functions and decentralized critics. Based on this decomposition, we develop Agent-Chained Policy Optimization (ACPO), where actors are trained independently, with their updates together constituting a single step on the joint policy gradient. Central to this result is a serialized view of the simultaneous joint decision in which agents commit actions one at a time, each conditioning on a belief over preceding actions that ties the independent per-agent updates into a single joint step. We evaluate on-policy and off-policy instantiations of ACPO on Multi-Robot Warehouse, SMACv2, and MA-MuJoCo, where it outperforms strong baselines, with the gap widening as the number of agents grows.
arXiv:2606.30182v2 Announce Type: replace
Abstract: AI models are rapidly improving at autonomous coding, as shown by benchmark progress and one-off demonstrations such as AI implementing a C compiler. However, existing coding benchmarks tend to focus on shorter tasks, and one-off demonstrations are hard to compare systematically because they often have some human guidance, and are not standardized or repeated across models. To address these challenges, we introduce MirrorCode, a long-horizon coding benchmark based on reimplementing entire software projects. In MirrorCode, AI agents must replicate the functionalities of an existing program, without access to its source code. AI solutions must match the original program's output exactly on end-to-end tests, including held-out tests. MirrorCode's 25 target programs span different areas of computing: Unix utilities, data serialization and query tools, bioinformatics, interpreters, static analysis, cryptography, and compression. Existing AI models can already reimplement complex software, with the strongest model scoring 56% across the benchmark. For example, AI can reimplement gotree, a 16,000-line bioinformatics toolkit - a task that we believe would take weeks for a human engineer. However, studying the frontier of performance requires a larger inference budget than typical benchmarks, for example, \$2,600 over 19 days for a single attempt on a large task. We show that AI agents can already complete long-horizon software engineering tasks, especially when requirements are precisely specified. More broadly, our work suggests AI will have transformative effects on software engineering, as autonomous agents continue to improve.
arXiv:2606.30670v2 Announce Type: replace
Abstract: Large spin polarization observed in chiral-induced spin selectivity (CISS) remains difficult to explain quantitatively. Experimental polarizations measured in chiral molecular systems are often substantially larger than expected from weak microscopic spin-orbit interaction in organic materials. We revisit the geometrical spin current introduced previously for electrons constrained to curved paths and propose a scale argument relevant to transport through double-stranded DNA (dsDNA). Estimation of the intrinsic geometrical spin-current scale using representative dsDNA parameters yields current magnitudes substantially larger than measured transport currents in many CISS experiments. We suggest that this hierarchy allows transport to be interpreted as weak leakage from underlying curved spin-current states. Within this picture, large CISS polarization emerges as transport selection of pre-existing spin-current branches.
arXiv:2607.15452v1 Announce Type: cross
Abstract: Research on Nash equilibrium existence for infinite games has grown into a patchwork of technical preconditions and counterexamples. This paper presents a unified program in equilibrium theory by revising the predominant model of mixed strategies based on countable additivity. A game is specified by a nonempty set of players and, for each player, a nonempty action set and a bounded von Neumann-Morgenstern utility function. Every such game is shown to admit a Nash equilibrium in finitely additive mixed strategies. In addition, the equilibrium correspondence for any such game is shown to be nonempty, compact-valued, and upper hemicontinuous, and the same is true for equilibria obtained as limits of finite approximations. Techniques developed in this paper show that infinite games long treated as intractable become amenable to direct equilibrium analysis.
arXiv:2607.15721v1 Announce Type: new
Abstract: Cardiometabolic diseases remain among the most persistent drivers of preventable morbidity because diabetes, hypertension, and cardiovascular disease frequently co-occur and share metabolic, vascular, demographic, and behavioral determinants. Existing machine learning studies for chronic disease prediction often emphasize discrimination on a single dataset, while underreporting label leakage, calibration, temporal robustness, external transportability, and subgroup reliability. This paper presents CardioMeta, a calibrated multi-task framework for joint prediction of diabetes, hypertension, and cardiovascular disease across population survey and electronic health record (EHR) data. The study uses NHANES for population-level model development and temporal validation, and MIMIC-IV for EHR-domain evaluation under substantial distribution shift. To reduce circular label reconstruction, the primary analysis excludes disease-defining variables from the corresponding prediction heads, while a full-clinical feature setting is retained only as sensitivity analysis. CardioMeta combines a shared cardiometabolic encoder with disease-specific gated heads and post-hoc probability calibration. In the leakage-reduced temporal validation setting, the model achieved a macro-AUROC of 0.839, macro-AUPRC of 0.536, macro-F1 of 0.614, and expected calibration error of 0.024, with modest but consistent improvements over strong gradient-boosting and neural tabular baselines. External evaluation on MIMIC-IV showed clear degradation under domain shift, while limited fine-tuning partially recovered performance. The findings indicate that the principal value of multi-task cardiometabolic modeling lies not in inflated accuracy, but in reproducible leakage control, calibrated probabilities, and transparent reliability reporting across heterogeneous healthcare data sources.
arXiv:2607.15969v1 Announce Type: new
Abstract: This paper presents a trajectory prediction method for marine vessels based on optimal planning. Crude initial trajectories respecting static obstacles are first generated using A*-search to provide a feasible warm start. In the second step, a numerical optimizer is used to ensure COLREG compliance. The prediction problem is posed as sequential trajectory planning from the perspective of each surrounding vessel, requiring only their current positions, velocities, and intended destinations as input. As the latter is included in AIS messages, this enables faster predictions than learning-based methods that typically require longer data histories. The proposed method is validated using real-world scenarios constructed from AIS data.
arXiv:2607.15415v1 Announce Type: new
Abstract: Memristive crossbar arrays have emerged as a promising platform for efficient on-chip learning, enabling local learning rules such as Equilibrium Propagation (EP) to be realized without the memory overhead of conventional backpropagation. However, as these devices age, permanent stuck-at (SA) faults accumulate at individual synaptic elements, irreversibly corrupting the stored weights and degrading model performance in ways that in-situ retraining alone cannot address. Existing mitigation strategies either rely on hardware redundancy at significant area and power cost, or require explicit fault localization that is impractical to perform continuously on-chip. This paper adopts a brain-inspired fault recovery route motivated by self-repair functionalities enabled by astrocytes -- a type of glial cell. We couple a computational neuroscience study of astrocytic neuromodulation under permanent synaptic faults with an algorithmic recovery mechanism for faulted crossbar-mapped networks. Our computational neuroscience study characterizes how astrocytes modulate surviving synapses in a bidirectional EP-specific network setting under both SA-0 and high-conductance stuck-at fault conditions, revealing that recovery operates at the neural population level. Motivated by this observation, we propose a retraining-based repair mechanism that augments EP with an additional repair nudge anchored to pre-fault activation targets of the healthy network, encouraging surviving weights to collectively reconstruct pre-fault internal representations without any information of which weights are faulty. To our knowledge, STAR is the first astrocyte-inspired repair mechanism to operate under a supervised local learning rule and to scale to convolutional networks.
arXiv:2607.02478v2 Announce Type: replace
Abstract: In this letter, we present a strategy for autonomous docking of autonomous vehicles in three-dimensional space. Docking is a safety-critical task and requires expert piloting skills. Vehicles with autonomous docking capabilities are highly desirable in various applications, such as marine vehicle docking, aerial vehicle docking, spacecraft docking, and landing. To dock autonomously with the docking station, the vehicle must align itself to a specific desired orientation relative to the docking station and also reduce speed as it approaches. The vehicle achieves near-zero speed to dock successfully and safely without colliding with the docking station. Inspired by the philosophies from the guidance literature, we present a finite-time sliding mode-based strategy to achieve the same. The range and line-of-sight kinematics relations describing the motion of the vehicle with respect to the stationary docking station are used to steer the vehicle to achieve the desired orientation for docking. This docking strategy is validated in MATLAB\textsuperscript{\textregistered} simulations for various initial locations and orientations of both the vehicle and the docking station.
arXiv:2607.02672v2 Announce Type: replace
Abstract: Local pairwise comparisons are a standard tool for learning how people want decision rules to work, e.g., in participatory design or alignment. However, their use builds in two strong assumptions: that local comparisons are sufficient evidence about how a person wants an automated decision rule to behave, and that people can always answer those comparisons decisively. We investigate how these assumptions may be compromised under internal pluralism: the idea that an individual evaluates decision rules according to multiple authoritative priorities about how the rule should behave. We provide a formal model of such pluralistic preferences over decision rules, which then lets us identify two distinct failures of forced local pairwise comparison data. First, priorities such as proportionality, egalitarianism, and equal treatment are inherently global: what they imply in one case can depend on what happens elsewhere, so local comparisons may fail to capture them. Second, even when priorities are representable locally, tension between strongly-held priorities can generate internal conflict, producing potentially costly behavioral distortions when comparisons are forced. We then use our model to investigate the alternative -- allowing people to report indecision -- and our findings suggest that doing so can considerably reduce the number of queries needed to learn preferences accurately. We conclude by describing how our model points toward preference-learning methods that elicit these priorities directly, yielding more faithful and interpretable accounts of what people value.
arXiv:2607.03651v2 Announce Type: replace
Abstract: While traditional hub capacity planning models optimize effectively for quantitative inputs, they often fail to digest qualitative business context. We propose a novel framework where a large language model (LLM) agent iteratively proposes hub capacity decisions guided by natural-language business context descriptions. The key mechanism is a chain-of-thought reasoning protocol: the LLM constructs a structured decision table that maps each contextual item to specific capacity adjustments based on the implied direction and magnitude of changes. The new capacity decision is then validated through a feedback loop with an optimization model, which provides routing-based performance metrics to guide the agent's selection. On a real-world 13-hub freight network in the southeastern US, our framework achieves a 2.8% optimality gap relative to the hidden ground-truth, a significant improvement over the 11.0% gap produced by the traditional optimization model without textual business inputs. This demonstrates that LLMs can serve as a contextual bridge, integrating qualitative business insights into Operations Research workflows.
arXiv:2607.03919v2 Announce Type: replace
Abstract: Polar codes are proven to be capacity-achieving codes, being gradually practiced in wireless communications. However, their successive cancellation (SC) and successive cancellation list (SCL) decoding incur latency challenge especially for long codes. This paper proposes the generalized integrated interleaved (GII)-polar codes for block Rayleigh fading channels, yielding both reduced decoding latency and competent decoding performance. Under the GII paradigm, two consecutive polar codewords of length $N$ are virtually coupled through a nested codebook which has a lower dimension, capable of correcting more errors. The component polar codes are known as interleaves of a GII-polar code. If decoding of an interleave fails, it can be projected into the decoding of the nested one, enabling richer error patterns to be corrected by each interleave. Since decoding of the interleaves can be performed in parallel, GII-polar codes yield a reduced decoding latency over a single polar code of length $2N$. Our simulation results validate both the decoding latency and performance merits of the proposed coding scheme under block Rayleigh fading channel.
arXiv:2607.04020v2 Announce Type: replace
Abstract: Uterine diseases represent an important category of gynecologic pathology and require accurate histopathological assessment for diagnosis and treatment planning. Whole-slide images (WSI) have enabled the digital transformation of pathology workflows and provided new opportunities for artificial intelligence (AI) in computational pathology. In particular, multimodal models that jointly analyze histopathology images and pathology reports have shown promising potential for automated pathology report generation and AI-assisted diagnosis. However, the development of such systems remains limited by the scarcity of datasets that pair whole-slide images with clinically meaningful pathology reports. Instead, existing pathology datasets focus on patch- or slide-level annotations of a single endpoint (e.g., disease class), which do not fully capture the rich information in full clinical diagnostic workflow reports. Here, we introduce TUM-Uteria, a uterine pathology dataset comprising WSIs paired with diagnostic pathology reports at both the case and slide levels, collected from a tertiary medical center. The dataset contains 216 clinical cases, comprising 455 slide-level WSI-report pairs. The dataset underwent a structured multi-stage validation procedure involving board-certified pathologists to ensure reliable annotations. TUM-Uteria supports research in computational pathology, including whole-slide image analysis, multimodal learning, and automated pathology report generation.
arXiv:2607.04419v4 Announce Type: replace
Abstract: Pooling, substituting, or reusing evaluator-derived step rewards assumes that their direction survives a change of evaluation channel. The same frozen transition can violate that assumption. Process rewards vary agent states, while evaluator audits vary scoring configurations; neither first difference isolates their interaction. We define Agent Step Value (ASV) as a channel-indexed target-margin gain and identify the state-by-channel interaction on complete matched faces. Across frozen PubMed question-answering transitions, direct scoring yields a positive mean ASV, while the generated-view channel yields a negative mean. Two matched replay waves reproduce this reversal, and cross-channel sign disagreement exceeds same-channel retry disagreement by 48.0 percentage points. Matched retrieval faces localize the reversal to the generated-view coordinate and trace its direction across a readout-and-stack bridge. A source-only generation contract restores the positive mean direction on artifact-bearing retrievals and removes parser-detected substantive support claims from artifact-free before-state views. ASV turns channel sensitivity into an identified measurement problem that can be localized and tested by intervention before step rewards are reused.
arXiv:2607.05356v2 Announce Type: replace
Abstract: Streaming 3D reconstruction relies on a compact recurrent scene state to process long image streams in linear time and bounded memory. However, repeated updates can gradually corrupt this state, causing reliable historical information to be overwritten by noisy or ambiguous observations. We introduce ReCal3R, a reliability-calibrated learning rate method for recurrent 3D reconstruction. Instead of directly applying a candidate learning rate, our method estimates state token reliability from the maintained scene state and uses it to calibrate a candidate learning rate derived from token alignment, state reconstruction residual, and recent update pressure. The resulting token-wise learning rate interpolates between a conservative base rate and the candidate rate, suppressing aggressive updates on unreliable tokens while preserving adaptation to informative frames. Applied to CUT3R as a training-free calibration rule, ReCal3R reaches strong performance on long sequences in pose, depth, and reconstruction quality, including a 3.7$\times$ reduction in ATE, with comparable runtime and memory. Code is available at: https://github.com/Powertony102/ReCal3R.