arXiv:2505.17423v4 Announce Type: replace
Abstract: Many decision-making tasks, where both accuracy and efficiency matter, still require human supervision. For example, tasks like traffic officers reviewing hour-long dashcam footage or researchers screening conference videos can benefit from concise summaries that reduce cognitive load and save time. Yet current vision-language models (VLMs) often produce verbose, redundant outputs that hinder task performance. Existing video caption evaluation depends on costly human annotations and overlooks the summaries' utility in downstream tasks. We address these gaps with Video-to-text Information Bottleneck Evaluation (VIBE), an annotation-free method that scores VLM outputs using two metrics: grounding (how well the summary aligns with visual content) and utility (how informative it is for the task). VIBE selects from randomly sampled VLM outputs by ranking them according to the two scores to support effective human decision-making. Human studies on LearningPaper24, SUTD-TrafficQA, and LongVideoBench show that summaries selected by VIBE consistently improve performance-boosting task accuracy by up to 61.23% and reducing response time by 75.77% compared to naive VLM summaries or raw video.
Science Journals
arXiv:2602.22599v2 Announce Type: replace
Abstract: We study with ab-initio theory high harmonic generation (HHG) from a series of gas-phase fullerenes (from C$_{20}$ to C$_{60}$, including isomers) under varying laser conditions (linearly polarized at various angles and different ellipticities). We explore the role of multiple orbitals in the dynamics and show that due to the tight energy level spacing in these systems (forming the equivalent of energy bands), multiple orbitals contribute and cause severe spectral interferences. HHG cutoff and plateau ranges are shown relatively similar across species of different sizes. We further explore the role of correlations, which are known to be prevalent in C$_{60}$ and similar species and responsible for plasmonic resonances. We find that the independent-particle approximation, where electrons are not dynamically interacting, fails to quantitatively describe features in the spectrum besides the cutoff energy. This failure is broad across the HHG plateau and perturbative regime in all species. Broadly, correlations are seen to reduce HHG yields and cause sharper emission peaks. Lastly, we study HHG ellipticity and angular dependence across species and show that: (i) HHG angular sensitivity becomes less prevalent for larger fullerenes. This arises from a more isotropic structure of the higher point-group molecules. (ii) HHG yields decay faster with the driving laser ellipticity for smaller systems, owing to their reduced size creating smaller recombination cross sections with re-colliding electron wave packets (with C$_{60}$ posing an anomalous exception to this rule). Our predictions pin-point fullerenes as an ideal system for exploring multi-electron interactions in HHG and strong-field physics, and should motivate experiments.
arXiv:2602.23565v2 Announce Type: replace
Abstract: In many economically relevant contexts where machine learning is deployed, multiple platforms obtain data from the same pool of users, each of whom selects the platform that best serves them. Prior work in this setting focuses exclusively on the "local" losses of learners on the distribution of data that they observe. We find that there exist instances where learners who use existing algorithms almost surely converge to models with arbitrarily poor global performance, even when models with low full-population loss exist. This happens through a feedback-induced mechanism, which we call the overspecialization trap: as learners optimize for users who already prefer them, they become less attractive to users outside this base, which further restricts the data they observe. Inspired by the recent use of knowledge distillation in modern ML, we propose an algorithm that allows learners to "probe" the predictions of peer models, enabling them to learn about users who do not select them. Our analysis characterizes when probing succeeds: this procedure converges almost surely to a stationary point with bounded full-population risk when probing sources are sufficiently informative, e.g., a known market leader or a majority of peers with good global performance. We verify our findings with semi-synthetic experiments on the MovieLens, Census, and Amazon Sentiment datasets.
arXiv:2602.23993v2 Announce Type: replace
Abstract: We present gradiend, an open-source Python package that operationalizes the GRADIEND method for learning feature directions from factual-counterfactual MLM and CLM gradients in language models. The package provides a unified workflow for feature-related data creation, training, evaluation, visualization, persistent model rewriting via controlled weight updates, and multi-feature comparison. We demonstrate gradiend through an English pronoun running example, a semantic sentiment use case that evaluates lexical generalization to held-out target words, and a large-scale feature comparison.
arXiv:2603.00759v3 Announce Type: replace
Abstract: In this paper, we present an online method for converting an arbitrary geometric path, represented by a sequence of states, and generated by any planner (e.g., sampling-based planners such as RRT or PRM, search-based planners such as ARA*, etc.), into a kinematically feasible, jerk-limited trajectory. The method generates a sequence of quintic/quartic splines that can be discretized at a user-specified control rate and streamed to a low-level robot controller. Our approach enables real-time adaptation to environmental changes and can be re-invoked at any instant to generate a new trajectory from the robot's current state to a desired target state or sequence of states. Under a bounded-obstacle-velocity assumption, the method provides conditional stopping-safety guarantees over a finite time interval in dynamic environments, while allowing bounded geometric deviation from the original path. Kinematic constraints, including jerk limits, are explicitly considered. We validate the approach in a comparative simulation study against a competing method, demonstrating favorable behavior w.r.t. smoothness, computational time, and real-time performance, particularly with frequent target-state changes (up to 1 [kHz]). Real-robot experiments demonstrate applicability in real-world scenarios, including scenarios with a human as an obstacle.
arXiv:2603.01202v2 Announce Type: replace
Abstract: This study examines how vulnerability is produced for human operators of Tesla's Full Self-Driving (FSD), a Level 2 semi-autonomous vehicle (SAV) system, by applying Florencia Luna's layered vulnerability framework. While existing road safety models conceptualize vulnerability as a fixed attribute of external road users, emerging evidence suggests that semi-autonomous vehicle operators themselves experience dynamic and situational vulnerability as they supervise automated systems that they do not fully control. To investigate this phenomenon, we conducted semi-structured interviews with 17 active FSD users, analyzing their accounts through a combined deductive-inductive coding process aligned with Luna's framework. Findings reveal three interacting layers of operator vulnerability, namely psychological, operational, and social. Vulnerability emerged not from any single layer but from how these layers converged in specific situations, creating fluctuating supervisory demands and uneven capacity to recognize and manage risk. The findings extend debates on contextual trust calibration, automation complacency, and meaningful human control by demonstrating how factors commonly treated as liabilities such as trust or informal learning, can both increase and mitigate vulnerability depending on context. This analysis determines the need for design and regulatory interventions that address psychological, operational, and social conditions together rather than in isolation, and highlights how responsibility is implicitly shifted onto individual operators within inadequately supported supervisory regimes.
arXiv:2607.09757v1 Announce Type: new
Abstract: Low-Rank Adaptation (LoRA) has become a cornerstone of parameter-efficient fine-tuning (PEFT); however, the conventional practice of uniform rank assignment ignores the functional heterogeneity of neural layers. Existing rank allocation methods typically struggle with a trade-off between computational intensity and heuristic simplicity: training-based methods suffer from prohibitive overhead, while pre-allocation methods fail to capture the dynamic task-specific representation manifold. In this paper, we propose RSLoRA (Representational Sensitivity LoRA), a training-free and gradient-free rank allocator driven by activation-space geometry. We identify a "sensitivity regime shift" across layers, observing that static weight analysis and local gradients are insufficient to reflect how updates reshape a model's internal representations. To address this, RSLoRA introduces a virtual representational probing mechanism. By simulating adaptation through structured low-rank noise and measuring the resulting manifold displacement by using Effective Rank and Frechet Distance, we identify high-sensitivity modules that require higher rank capacity. Our framework effectively bridges the gap between expert-crafted heuristics and actual representational impact. Extensive evaluations demonstrate that RSLoRA consistently outperforms state-of-the-art allocators (e.g., AdaLoRA, GoRA) across mainstream benchmarks. By eliminating the need for iterative training-time adjustments and backward gradients, RSLoRA provides a highly efficient, robust, and representation-aware solution for large-scale model adaptation.
arXiv:2607.10731v1 Announce Type: new
Abstract: We consider a variant of the bin packing problem with constraints on the number of copies of each item and their placement in the packing. The input $D_q := DD\ldots$ is defined as $q$ consecutive copies of the multiset $D$, with a fixed bin capacity $S$. Note that, for each item in $D$, there are $q$ copies in $D_q$. The goal is to pack all the items in $D_q$ into the minimum number of bins, such that each bin contains at most one copy of each item and the total size of all items in a bin does not exceed the bin capacity $S$. We call this problem $q$BP. First Fit Decreasing ($\mathsf{FFD}$) is a classical bin packing algorithm: it first orders the items in nonincreasing order, then packs the next item into the first bin where it fits. In the literature, $\mathsf{FFD}$ proofs rely on the assumption that the last bin in the $\mathsf{FFD}$ packing contains only a single item. This assumption does not naturally extend to the $q$BP problem. In this paper, we circumvent this difficulty by analyzing $\mathsf{FFDq(D_q)}$ on a carefully chosen subinstance ${D'}_q \subseteq D_q$ ($q$ consecutive copies of $D$, each copy sorted in non-increasing order) while preserving the same upper bound for the original input $D_q$. We show that the approximation ratio of $\mathsf{FFDq(D_q)}$ for some special cases is
\begin{align*}
\mathsf{FFDq(D_q)} \leq \frac{11}{9}\mathsf{OPT(D_q)} + 3q
\end{align*} where $\mathsf{FFDq}$ and $\mathsf{OPT}$ denote the number of bins used by the $\mathsf{FFD}$ generalization and by an optimal algorithm, respectively.
arXiv:2607.10489v1 Announce Type: new
Abstract: We introduce Grassmannian splatting, a dynamic scene representation whose primitives are Gaussians supported on 3-planes in spacetime $\R^4$: generically, spatial 2-planes in uniform translation along their normals. Each primitive carries a unit normal $n \in \mathbb S^3/\{\pm 1\} \cong \mathrm{Gr}(3,4)$ and an unconstrained factor $L \in \mathbb R^{4 \times 3}$, with covariance \[
\Sigma_{4\mathrm{D}} = (P_n L)(P_n L)^T, \qquad P_n = I - n n^T. \] For generic $L$ and $n \neq \pm e_0$, conditioning on time returns a rank-2 surfel at every frame. The normal of the disk and its velocity along that normal are read off from $n$; the disk shape and the tangential drift of its center are set by $L$. Existing native 4D Gaussian splatting methods [\it{Yang et. al. 2023,Duan et. al. 2024}] slice full-rank spacetime covariances, so their per-frame primitive is a volumetric ellipsoid; since conditioning lowers rank by exactly one, a rank-2 surfel in the slice requires a rank-3 spacetime covariance, and the parameterization above realizes exactly these. The motion model is closed form, i.e. no deformation field is learned, and no custom CUDA is required: the conditioned disk feeds a standard 3DGS rasterizer through its precomputed-covariance interface. A soft clamp in the Schur denominator regularizes the static orientation and continuously bridges rank-3 static and rank-2 dynamic behavior, so static and moving primitives form a single continuous family. On the 17 HyperNeRF scenes of MonoDyGauBench, training is fastest among all compared methods (4.9 to 5.6 times faster than the strongest quality baselines), while ranking second in PSNR, MS-SSIM, and LPIPS. Code: https://github.com/PaulCelanCoding/grassmannian-splatting
arXiv:2607.11551v1 Announce Type: new
Abstract: We study trajectory prediction from libraries of stored output templates. Given the past of an unknown trajectory, the goal is to predict its future without identifying the state-space model that generated it. We show that libraries of trajectories generated by one or more dynamical systems define behavioral spaces that can be used as prediction mechanisms. For linear systems, we characterize exact prediction in terms of continuation maps, behavioral containment, and spectral conditions on output-visible eigenvalues. We also analyze robustness to noisy observations and noisy libraries, derive error bounds for out-of-library trajectories, and show how interconnection constraints can compose template libraries into new behavioral spaces with emergent modes. Finally, we extend the framework to nonlinear systems whose output trajectories are contained in, or immersed into, finite-dimensional linear behaviors. These results provide a theory of template-based prediction machines capable of generalizing beyond the stored trajectories and, in some cases, beyond the systems that generated them.
arXiv:2607.10806v1 Announce Type: new
Abstract: Quantifying abstractiveness in generated summaries is essential for evaluating summarization models beyond
surface-level metrics like ROUGE. We introduce Reference Abstraction (RA), Summary Abstraction (SA), and Abstraction
Ratio (AR) -- a set of principled heuristic metrics that measure how much a summary diverges from extractive copying
of the source text. The formulation uses the harmonic mean of document lengths modulated by a cubic non-overlap
factor, yielding dimensionally consistent, bounded output with non-linear sensitivity to the extractive-abstractive
boundary. Evaluation on 100 XSUM documents across four summarization models (BART-large-cnn, Pegasus-xsum, DistilBart,
MT5-small) demonstrates that the metrics successfully discriminate between extractive models (SA ~ 0.12-0.26) and
abstractive models (SA ~ 0.96-1.77), and that the Abstraction Ratio identifies summaries requiring manual evaluation
for potential hallucination. Code and results are available at https://github.com/katweNLP/AbstractionStudy.
arXiv:2209.12819v4 Announce Type: replace
Abstract: In the Maker-Breaker positional game, Maker and Breaker take turns picking vertices of a hypergraph $H$, and Maker wins if and only if she possesses all the vertices of some edge of $H$. Deciding the outcome (i.e., which player has a winning strategy) is a PSPACE-complete problem even when restricted to 4-uniform hypergraphs [Galliot, 2025]. As for hypergraphs of rank 3, the linear subcase (i.e., any two distinct edges intersect on at most one vertex) has been solved by obtaining a structural characterization of the outcome and a polynomial-time algorithm to decide it [Kutz, 2005]. A conjecture [Rahman and Watson, 2020] implies that the same results can be obtained for general hypergraphs of rank 3, which we confirm in this paper. We provide a structural characterization of the outcome and a description of both players' optimal strategies, all based on intersections of some key subhypergraph collections. From this, we derive a polynomial-time algorithm, thus closing the complexity gap for Maker-Breaker games relative to the size of the edges. Another corollary of our structural result is that, if Maker has a winning strategy on a hypergraph of rank 3, then she can ensure to win the game within a number of rounds that is logarithmic in the number of vertices.
arXiv:2505.22071v4 Announce Type: replace
Abstract: This work focuses on the end-to-end forecast of global extreme marine heatwaves (MHWs), which are unusually warm sea surface temperature events with profound impacts on marine ecosystems. Accurate prediction of extreme MHWs has significant scientific and financial worth. However, existing methods still have certain limitations in forecasting general patterns and extreme events. In this study, to address these issues, based on the physical nature of MHWs, we created a novel hybrid data-driven and numerical MHWs forecast framework Ocean-E2E, which is capable of 40-day accurate MHW forecasting with end-to-end data assimilation. Our framework significantly improves the forecast ability of MHWs by explicitly modeling the effect of oceanic mesoscale advection and air-sea interaction based on a dynamic kernel. Furthermore, Ocean-E2E is capable of end-to-end MHWs forecast and regional high-resolution prediction, allowing our framework to operate completely independently of numerical models while outperforming the current state-of-the-art ocean numerical/AI forecasting-assimilation models. Experimental results show that the proposed framework performs excellently on global-to-regional scales and short-to-long-term forecasts, especially in those most extreme MHWs. Overall, our model provides a framework for forecasting and understanding MHWs and other climate extremes. The source code is available at https://github.com/ChiyodaMomo01/Ocean-E2E.
MVMGNN;Multi-View Masked Graph Neural Network for Alzheimer's Disease Diagnosis using Structural MRI
arXiv:2607.09788v1 Announce Type: new
Abstract: Alzheimer's disease (AD) is a common neurodegenerative disorder, and early diagnosis is of great significance for delaying disease progression and enabling timely intervention. Mild cognitive impairment (MCI), which represents an intermediate clinical stage between cognitively normal aging and AD. Structural magnetic resonance imaging (sMRI) provides detailed characterization of anatomical structures and plays an important role in AD-related brain analysis. However, existing sMRI-based brain network methods typically rely on a single graph construction strategy, limiting their ability to jointly capture spatial relationships and morphological similarities between brain regions. To address these issues, this paper proposes an sMRI-based multi-view masked graph neural network model (MVMGNN) for AD diagnosis. A joint node-edge masking mechanism is proposed to simultaneously select radiomics feature dimensions and structural connections, reducing redundancy during graph learning. Furthermore, a patient-level cross-view gated fusion mechanism is proposed to integrate multi-view representations. Experimental results on the ADNI dataset demonstrate that MVMGNN outperforms several competing approaches in AD classification. Interpretability analysis further demonstrates that MVMGNN is able to identify key brain regions associated with AD, providing useful insights into discriminative patterns in sMRI-based brain networks.Our implementation is publicly available at https://github.com/chenzhao2023/MVMGNN_AD
arXiv:2607.09792v1 Announce Type: new
Abstract: Navigation is a fundamental capability of autonomous systems, yet most existing approaches rely on highly structured models and strong prior assumptions, limiting their robustness in open and uncertain real-world environments. Vision-and-Language Navigation (VLN) offers a promising direction by enabling robots to integrate natural language understanding with visual perception in a data-driven manner. Although VLN has attracted increasing research attention, systematic methodological taxonomy and real-world validation remain limited. This survey presents a comprehensive review of VLN research. Specifically, state-of-the-art methods are organized along two orthogonal dimensions: action paradigms, including hierarchical and monolithic frameworks, and model paradigms, including discriminative and generative approaches. A critical analysis of their respective strengths and limitations is provided. Additionally, we conduct a systematic real-world evaluation of representative VLN system configurations on a physical robotic platform. Experiments across ten diverse real-world scenes show a substantial performance gap between simulation and real-world deployment under the tested configurations: a representative monolithic RGB-only method achieves 61% success in simulation but drops to 22% in real-world deployment, while a hierarchical framework achieves a higher real-world success rate of 51%, suggesting stronger robustness in our evaluation setting. Finally, we highlight key challenges in perception, decision-making, and control that must be addressed in future research.
arXiv:2607.11053v1 Announce Type: new
Abstract: Fruitful collaborations rely on cooperative communications, including of contextual cues to incorporate into reasoning. The increasing use of LLMs in collaborative and agentic pipelines raises questions about the extent to which they exhibit these pragmatic capabilities, especially in scenarios where they may not have access to the same information as their collaborators. In this paper, we perform a novel investigation into the pragmatic reasoning capabilities of LLMs in a multi-party collaborative task under partial information conditions. We formalize a notion of collaborative epistemic asymmetry that explicitly connects objective task success to Grice's cooperative principle and empirically assess various LLMs' abilities to act cooperatively as both speakers and listeners, including both prompting and post-training strategies. Our results show that while LLMs exhibit certain pragmatic capabilities in collaborative settings, and these can be elicited through prompting and post-training, they still face challenges in pragmatic communication with incomplete information, and that certain failure modes do correlate with floutings of Grice's maxims that go unrecognized.
arXiv:2607.11683v1 Announce Type: new
Abstract: Graph retrieval-augmented generation (GraphRAG) enhances large language models with structured knowledge, yet existing systems construct knowledge graphs in a single extraction pass, producing noisy entities and brittle retrieval. RAGU, an open-source modular GraphRAG engine, addresses this by separating extraction from consolidation: entities and relations pass through two-stage typed extraction, DBSCAN-backed deduplication, LLM summarization, and Leiden community detection. A key insight motivates a compact extractor: the skills an in-pipeline LLM needs - comprehension, extraction, reasoning over context - are language skills that grow only weakly with model size, unlike factual world knowledge. Accordingly, we train Meno-Lite-0.1, a 7B model optimized for language skills, which outperforms Qwen2.5-32B on knowledge-graph construction (+12.5% relative harmonic mean) and matches it on English GraphRAG tasks. On GraphRAG-Bench (Medical), RAGU retrieves the most complete context at every factoid level (evidence recall up to 0.84 vs. $\leq$0.76) and overtakes HippoRAG2 on synthesis tasks; on multi-hop factoid QA, the apparent HippoRAG2 advantage is shown to be largely an answer-format artifact. RAGU is installable via $\texttt{pip install graph_ragu}$, runs on a single GPU, and is released under MIT. The source code is publicly available at https://github.com/RaguTeam/RAGU, and the Meno-Lite-0.1 model can be obtained from https://huggingface.co/bond005/meno-lite-0.1.
arXiv:2607.11481v1 Announce Type: new
Abstract: Humans routinely wield tools, swap grasps, and reposition objects within a single hand, seamlessly orchestrating contact transitions that span translation, reorientation, and finger gaiting. Endowing robot dexterous hands with this level of in-hand dexterity through teleoperation requires precise control of object motion via dynamic hand-object contact, yet current teleoperation systems remain far from this capability. To bridge this gap, we take a major step towards human-level dexterous teleoperation by introducing TeleDexter, a hand-object co-tracking controller that maps operator intent into learned, low-level contact execution. The controller is trained on consecutive co-tracking subgoals derived from human reference motions, utilizing a hybrid reward that couples sparse subgoal objectives with dense tracking rewards to enable learning across diverse interaction modalities rather than frame-wise trajectory imitation. The entire pipeline requires only single-stage RL and, with random action masking and domain randomization, transfers zero-shot to the real robot. We evaluate TeleDexter on seven challenging dexterous teleoperation tasks spanning object reorientation and long-horizon tool use across two dexterous hands, achieving a 75% average success rate where all baselines consistently fail. Furthermore, the collected demonstrations successfully train autonomous policies via behavioral cloning, marking a concrete step towards human-level dexterous teleoperation.
arXiv:2607.10140v1 Announce Type: new
Abstract: Existing optical flow methods broadly follow two paradigms: iterative optimization and diffusion-based estimation. Iterative methods, exemplified by RAFT, achieve high accuracy through recurrent refinement, but remain challenged by large displacements and complex motion. Diffusion-based methods introduce generative modeling and show promise in such ambiguous regions. However, existing diffusion models usually denoise the entire dense flow field from Gaussian noise, including simple regions where reliable motion can already be estimated by a lightweight network. This increases the denoising burden and may cause slow convergence and unstable training. To address this issue, we introduce FlowPainter, a diffusion-based optical flow framework that reformulates dense-flow generation as confidence-guided soft inpainting. FlowPainter employs a lightweight confidence-aware network to predict a rough flow and a pixel-wise confidence mask, distinguishing reliable simple regions from uncertain hard regions. The resulting simple-flow prior is used for confidence-based initialization and further injected into iterative denoising through confidence-gated residual guidance. With dynamically decaying guidance strength, FlowPainter stabilizes early denoising while preserving the flexibility of the diffusion model for late-stage detail refinement. Extensive experiments on public benchmarks, including Sintel, KITTI, and Spring, show that FlowPainter achieves strong accuracy under comparable training settings and converges more efficiently than existing diffusion-based optical flow methods, with notable gains on challenging benchmark splits. Our approach offers a practical way to integrate reliable discriminative priors with diffusion-based refinement for optical flow estimation. Our code is publicly available at https://github.com/mya012/FlowPainter.
arXiv:2607.10747v1 Announce Type: new
Abstract: Digital credential ecosystems increasingly combine multiple standards. Because implementations have evolved independently across jurisdictions and application domains, systems described under the common label ``digital credential'' often remain mutually non-interoperable. Conventional element-by-element comparisons of identifiers, data models, credential formats, protocols, and signature algorithms do not explain why interoperability fails even when stacks share a data model, nor do they identify what a verifier must obtain, and what it must trust, before accepting a credential. We present a verifier-centric conceptual model built on two decompositions. The first separates credential processing into signature verification (L1), semantic interpretation (L2), and validation (L3), and models the supporting materials through two orthogonal planes: Constitution, which captures ecosystem-level arrangements and trust declarations, and Logistics, which captures how verification materials are stored and delivered; the Shinken framework makes trust assumptions explicit across all five functions. The second characterizes where each function may be placed along three dimensions (placement, timing, and disclosure). From the condition of being verifiable, the model derives seven consequences, distinguished as definitional corollaries, operational implications, and design trade-offs. Applying the model to four learner-credential stacks and to existing ecosystems including authentication federations, we show that it explains interoperability failures, verifier-side burden, offline verifiability, privacy implications, and terminological ambiguities that element-wise comparison leaves unresolved.
arXiv:2603.04419v2 Announce Type: replace
Abstract: We characterize the phenomenon of context-dependent affordance computation in vision-language models (VLMs). Our primary study uses Qwen3-VL-30B-A3B ($n = 3{,}213$ scene-context pairs from COCO-2017: 479 images under 7 agentic personas), with a cross-model replication on LLaVA-1.5-13B. We demonstrate substantial affordance drift: mean Jaccard similarity between context conditions is $0.095$ (95% CI $[0.092, 0.097]$ across $N = 479$ images; $9{,}244$ prime pairs; $p < 0.0001$), indicating that more than 90% of lexical scene description is context-dependent; the LLaVA replication reproduces the effect (mean $J = 0.160$, 84% context-dependent). Sentence-level cosine similarity confirms drift at the semantic level (mean $= 0.415$, 58.5% context-dependent). Stochastic baseline experiments ($2{,}384$ inference runs across 4 temperatures and 5 seeds) confirm this reflects genuine context effects rather than generation noise: within-prime variance is substantially lower than cross-prime variance across all conditions. Tucker decomposition with bootstrap stability analysis ($n = 1{,}000$ resamples) reveals stable orthogonal latent factors: a "Culinary Manifold" isolated to chef contexts and an "Access Axis" spanning child-mobility contrasts. The gap between lexical (90%) and semantic (58.5%) measures indicates that surface vocabulary changes more than underlying meaning under context shifts. These findings suggest a direction for robotics: dynamic, query-dependent ontological projection (JIT Ontology) rather than static world modeling. We do not claim to establish processing order or architectural primacy; such claims require internal representational analysis beyond output behavior.
arXiv:2603.05559v2 Announce Type: replace
Abstract: Decision makers exploiting photonic chaotic dynamics obtained by semiconductor lasers provide an ultrafast approach to solving multi-armed bandit problems by using a temporal optical signal as the driving source for sequential decisions. In such systems, the sampling interval of the chaotic waveform shapes the temporal correlation of the resulting time series, and experiments have reported that decision accuracy depends strongly on this autocorrelation property. However, it remains unclear whether the benefit of autocorrelation can be explained by a minimal mathematical model. Here, we analyze a stochastic-process model for solving the two-armed bandit problem based on time series, where the threshold and a two-valued Markov signal evolve jointly. Numerical results reveal an environment-dependent structure: negative (positive) autocorrelation is optimal in reward-rich (reward-poor) environments. These findings show that negative autocorrelation of the time series is advantageous when the sum of the winning probabilities is more than one, whereas positive autocorrelation is useful when the sum of the winning probabilities is less than one. Moreover, the performance is independent of autocorrelation if the sum of the winning probabilities equals one, which is mathematically clarified. This study paves the way for solving the two-armed bandit problems for reinforcement learning applications in wireless communications and robotics.
arXiv:2603.05576v3 Announce Type: replace
Abstract: Generalizing skill policies to novel conditions remains a key challenge in robot learning. Imitation learning methods, while data-efficient, are largely confined to the training region and consistently fail on input data outside it, leading to unpredictable policy failures. Alternatively, transfer learning approaches offer methods for trajectory generation robust to both changes in environment and tasks, but they remain data-hungry and lack accuracy in zero-shot generalization. We address these challenges in the context of task inversion learning and propose a novel joint learning approach to achieve accurate and efficient knowledge transfer. Our method constructs a common representation of the forward and inverse tasks, and leverages auxiliary forward demonstrations from novel configurations to successfully execute the corresponding inverse tasks, without any direct supervision. We demonstrate the extrapolation capabilities of our framework through ablation studies and experiments in simulated and real-world environments that require complex manipulation skills with a diverse set of objects and tools, where we outperform diffusion-based and multimodal VAE alternatives.
arXiv:2603.09714v2 Announce Type: replace
Abstract: While multi-audio understanding is critical for large audio-language models (LALMs), it remains underexplored. We introduce MUGEN, a comprehensive benchmark evaluating this capability across speech, general audio, and music. Our experiments reveal consistent weaknesses in multi-audio settings, and performance degrades sharply as the number of concurrent audio inputs increases, identifying input scaling as a fundamental bottleneck. We further investigate training-free strategies and observe that Audio-Permutational Self-Consistency, which diversifies the order of audio candidates, helps models form more robust aggregated predictions, yielding up to 6.28% accuracy gains. Combining this permutation strategy with Chain-of-Thought further improves performance to 6.74%. These results expose blind spots in current LALMs and provide a foundation for evaluating complex auditory comprehension.
arXiv:2607.10895v1 Announce Type: new
Abstract: With the transition to remote instruction, new modalities for conducting hands-on labs are needed. In particular, courses that entail major hardware components faced challenges in making the hardware available for students in a reliable and sustainable way. This paper presents our experience in implementing a virtual breadboard feature to interface with real Field Programmable Gate Arrays (FPGAs) boards located on the University of Washington's campus, where students can access the boards remotely to complete their lab assignments.