Papers

Partial deepfake speech, where only limited segments of an utterance are synthesized or manipulated, poses a significant challenge to existing deepfake detection systems. As the proportion of spoofed regions decreases, passive detectors become increasingly unreliable, and accurate detection and restoration remain challenging. In this paper, we revisit audio steganography from a new perspective and propose its use as a proactive defense against partially deepfaked audio. In particular, we consider a self-embedding strategy in which a clean speech signal embeds a compressed representation of itself, enabling post-hoc extraction of reference content. We demonstrate how existing audio steganography methods can be repurposed to support detection of partial deepfakes through codec-based restoration. Experiments on a benchmark dataset show that the proposed approach complements passive defenses. Remarkably, the proposed method operates without any training, providing a robust and data-efficient alternative for partial deepfake detection.

Physical human-robot collaboration requires a robot to provide assistance when human intention is clear while remaining compliant when several future motions are plausible. We present an adaptive stiffness framework based on generative action-chunk sampling. Conditioned on an RGB image and external joint-torque estimates, the policy samples multiple latent variables from an observation-conditioned prior and decodes them into future action chunks. Variation among the sampled action chunks is used to continuously adapt joint stiffness and damping. Greater variation makes the robot more compliant to facilitate human guidance, whereas lower variation provides firmer assistance. In a real-world collaborative transport task with four possible directions, the proposed method achieved an average success rate of 0.95, compared with 0.83 for a fixed-stiffness ablation and 0.69 for a deterministic baseline. Near direction determination, variation among the sampled action chunks increased, and the controller reduced stiffness accordingly. These results suggest that variation among actions sampled by a generative policy can serve as an online control signal for balancing assistance and compliance in physical human-robot interaction.

Mixed-Integer Linear Programming (MILP) is a fundamental optimization paradigm in combinatorial optimization and has been widely applied across real-world domains. Due to its NP-hard nature, obtaining optimal solutions for large-scale or highly constrained MILP instances remains computationally prohibitive. Learning-based solution prediction has therefore emerged as a promising approach to provide high-quality variable assignment for solver acceleration. However, existing methods typically adopt a one-shot prediction paradigm that predicts the marginal probabilities of all variables simultaneously. As a result, the conditional dependencies among variables are only implicitly captured through message passing, with the burden of modeling the combinatorial structure falling entirely on the representational capacity of graph neural networks. To address this limitation, we propose the Structure-Aware Hierarchical Solution Prediction (SHSP) framework that replaces the parallel marginal decoding of one-shot methods with a novel hierarchical conditional decoding mechanism. Specifically, SHSP constructs a variable coupling graph from the constraint structure, decodes variables sequentially along a hierarchy of increasing coupling strength, and conditions each hierarchy on previously predicted assignments. To mitigate error accumulation during the decoding process, SHSP further incorporates a confidence-aware mask-and-repair mechanism to identify and correct unreliable intermediate predictions. We integrate SHSP with multiple learning-guided search methods, and evaluate it on four standard MILP benchmarks. Experimental results demonstrate that SHSP significantly outperforms existing one-shot prediction baselines, achieving a 54% average reduction in solution gap.

Paradigmatic interaction models explain how collective behaviors can emerge in complex systems from interactions among the constituent agents. In bio-inspired swarms, however, interactions alone may not suffice to bring the population to a desired aggregate configuration within a prescribed time horizon, as needed in applications ranging from targeted therapy to collective transport and emergency evacuation. In the present work, we consider finite-horizon minimum-energy collective steering for inertial swarms that are subject to stochastic disturbances. We focus on the mean-field representations of these multi-agent systems driven by Cucker–Smale alignment or Morse attraction–repulsion interactions. Our objective is to steer the swarm between prescribed endpoint distributions using a state-feedback control, where the endpoint specifications can be full phase-space distributions (positions and velocities) or position marginals alone. Our formalism is rooted in the theory of Schrödinger bridges, which has inspired contemporary developments spanning statistical inference, biological modeling, stochastic control, and generative learning. Within the bridges framework, the uncontrolled interacting stochastic dynamics are viewed as a prior model, and the optimal control as the minimum-energy corrective drift needed to realize the prescribed distributions. We derive nonlinear, coupled necessary optimality systems with a time-symmetric structure reminiscent of classical Schrödinger bridges, and propose nested fixed-point schemes to numerically solve them. Numerical examples show that the obtained optimal control (corrective drift) can dynamically exploit or counteract the interaction forces, depending on whether the latter are favorable or adversarial to the steering task.

The OBABO and BAOAB schemes and the other standard Strang splittings of kinetic (underdamped) Langevin dynamics are widely used Markov chain Monte Carlo algorithms. Under a suitable friction scaling, the underlying diffusion relaxes on a ballistic time scale, suggesting that these discretizations, suitably tuned, sample targets with condition number $\kappa$ in $O(\sqrt{\kappa})$ iterations. We prove that no fixed choice of step size and friction, based only on the curvature bounds and the dimension, achieves this acceleration: total variation mixing time lower bounds for OBABO show that ballistic cold-start mixing fails uniformly over the smooth strongly convex class, and the lower bounds extend, with the same orders, to BAOAB and the other four Strang splittings. The proof transfers non-acceleration from optimization to sampling. Eliminating velocity gives an exact noisy heavy-ball recursion, and by the non-acceleration theorem of Goujaud, Taylor and Dieuleveut, for every tuning either some Gaussian target has a mode with relaxation time at least of order $\kappa$, or an attracting cycle exists on a smooth potential; dilating such a potential as $U_R(x)=R^2U(x/R)$ preserves its curvature bounds and produces metastability for a number of steps exponential in $R^2$, from an initial state at Wasserstein distance $O(R)$ from equilibrium. Using contraction estimates of Leimkuhler, Paulin and Whalley and a Wasserstein-to-total-variation regularization estimate, we prove a complementary upper bound of $O(\kappa)$ steps, up to logarithmic factors, for a fixed-parameter OBABO tuning. Hence, among fixed-parameter OBABO tunings, the optimal condition-number dependence of cold-start total variation mixing over this class is linear, up to logarithmic factors. A direct Gaussian calculation also rules out fixed-parameter acceleration for the left-endpoint exponential integrator.

Multi-agent LLM systems coordinate through natural-language messages that consume 40–60% of their token budget. Replacing these with structured graphs reduces cost but fails on tasks requiring adaptive reasoning. We propose Routed Graph Handoff, where a lightweight LLM router (155 tokens, 0.15% overhead) selects between a typed dependency graph and natural language for each delegation. On four benchmarks (1,050+ trajectories), the routed system matches or exceeds NL-only on every task: +12.7\,pp on $\tau$-retail at 3.2$\times$ compression ($p{<}0.01$), +8.7\,pp on BrowseComp at 2.2$\times$ compression ($p{<}0.05$), and parity on BFCL and AppWorld. Without the router, graph-only delegation regresses 14.6\,pp on AppWorld; the router eliminates this at near-zero cost. A graph-aware executor prompt is required: the same schema without interpretation guidance yields no gain. An oracle analysis reveals 8.6\,pp of additional headroom, motivating execution-time adaptive routing as future work.

Mixture-of-Experts (MoE) architectures enable scalable and efficient large language models (LLMs) by selectively activating expert sub-networks through a routing mechanism. However, this adaptive design introduces a new attack surface: specific experts become disproportionately correlated with certain tokens (e.g., end-of-sequence), allowing adversaries to manipulate model behavior via lightweight perturbations. In this work, we present Groundhog Bit-Flip Attack (GBFA), the first bit-flip-based Denial-of-Wallet availability attack against MoE-based LLMs. By identifying and flipping routing-layer bits associated with related expert activations, we demonstrate that GBFA substantially extends the decoding token usage across three different LLM modes: conversational, reasoning, and agentic tasks, while largely preserving semantic fidelity. Across four main real-world MoE-based LLMs, manually deactivating on average fewer than 4 experts drives average output inflation to $\mathbf{5912\%}$, with the majority of test samples reaching max tokens. These results reveal a robustness vulnerability of MoE architectures to bit flip, and highlight the potential of GBFA as an availability attack against LLMs.

Learned traffic-behavior models are commonly trained separately for each intersection, creating model portfolios that cannot share evidence across sites. We present PhaseShift, a topology-aware framework that harmonizes heterogeneous roadside trajectories into a shared actor-centric representation and trains one reusable backbone. Ego-relative coordinates, trajectory-induced movement paths, normalized signal context, and variable-cardinality interaction tokens remove site conventions while preserving behaviorally relevant topology. The backbone supports pooled operation, zero-shot at a held-out intersection, and low-data adaptation. We evaluate five intersections in two Florida regions on balanced field data, 100k training windows and equal-sized test sets per site under a replay-conditioned, best-of-sampled-trajectory protocol. At 10s, one pooled model lowers both minADE and minFDE relative to trained local models at all five sites, with median reductions of 36.8% and 22.0%. Leave-one-intersection-out deployment, including one cross-region fold, beats local training on both 10-s metrics at four of five sites, although short-horizon performance is less uniform. Fine-tuning with 1,000 target update windows improves on zero-shot at three sites and is the strongest regime at one. At site 7, every cross-site mixture sharply lowers long-horizon error under a fixed 100k-window budget; test-likelihood gains argue against a best-of-sample dispersion-only explanation. Local models fall behind calibrated IDM at the two highest-flow sites after long autoregressive rollouts; pretrained-backbone regimes do not. Within this five-site evaluation, PhaseShift demonstrates consolidation across heterogeneous physical control settings while identifying sites that still require adaptation. The protocol measures conditional single-vehicle generation under replayed context, not closed-loop traffic simulation.

Fine-grained Cross-View Localization (CVL) estimates the precise position and orientation of a ground-level image by aligning it with geo-referenced aerial imagery, offering a scalable alternative to Global Navigation Satellite Systems (GNSS) in challenging urban environments. Existing datasets rely on data collected with high-end sensor suites, which inherently limit image diversity and scalability. While in-the-wild images are abundant, their noisy geo-tags make them unsuitable for reliable evaluation. To bridge this gap, we introduce OpenCVL, a large-scale, diverse, and open dataset containing 617,388 ground-aerial image pairs spanning 41 cities across four European countries. All images are sourced from permissive platforms, ensuring long-term accessibility and supporting open and reproducible research. The training set combines images captured with high-end sensors with diverse in-the-wild imagery. We further develop a data curation framework that filters and corrects pose annotations to construct reliable in-the-wild evaluation data. In addition, OpenCVL includes dedicated cross-area and snowy test sets to assess generalization and robustness. Experiments with a state-of-the-art CVL model on OpenCVL show that incorporating noisy in-the-wild data consistently improves performance on clean test sets, suggesting a promising direction for scaling CVL with diverse real-world imagery.

We analyze exact-metric, Metropolis-adjusted Dikin walks by keeping the proposal determinant and reverse quadratic form together. Their leading uncentered terms cancel in the complete logarithmic acceptance ratio, leaving centered fluctuations that can be controlled with second-order tools. For a polytope given by $n$ inequalities and a convex $L$-Lipschitz potential, this yields warm-start mixing in $\widetilde O((d^{2}+dL^{2}R^{2})\log(w/\delta))$ steps for the regularized Lee–Sidford walk. For a spectrahedron with $n\times n$ blocks, the log-det walk mixes in $\widetilde O((\psi^\star nd+dL^{2}R^{2})\log(w/\delta))$ steps, where $\psi^\star$ measures matrix leverage. The two analyses share an acceptance-to-mixing reduction. A proposal-comparison argument transfers the polytope bound to an appropriately padded $O(1/d)$-accurate metric computed from high-precision Lewis weights. For spectrahedra, given $\widehat\psi\ge\psi^\star$, a direct-or-two-seed TensorSRHT construction gives an exact-arithmetic implementation with $\psi^\star$ replaced by $\widehat\psi$ in the mixing bound.

Proof of reserves (PoR) can improve transparency for tokenized assets, but aggregate reserve coverage does not establish whether off-chain assets are legally eligible, unencumbered, consistently valued, or sufficiently liquid for redemptions. We propose RWA-PoB, a credential-based proof-of-backing framework for tokenized U.S. Treasury products. Five authorised institutional roles approve a canonical EIP-712 snapshot containing reserve, liability, liquidity, and policy information. The framework evaluates backing adequacy through the Backing Coverage Ratio (BCR) and short-term redemption capacity through the Redemption Liquidity Coverage (RLC). The Solidity prototype couples the policy controller to an ERC-20 token. Successful issuance atomically increases token supply and recorded liabilities by the corresponding USD-denominated liability. A redemption request burns tokens while reclassifying the corresponding obligation as pending. The obligation is reduced only after an authorised settlement-role account confirms payment. We evaluate the framework against a simplified aggregate PoR baseline using USDY-calibrated liabilities and deterministic synthetic reserve scenarios. Both approaches permit issuance in the valid state, but RWA-PoB rejects an encumbered-assets state with a BCR of 96.3408%, below the experimental 105% threshold. Under liquidity stress, it classifies the proposed redemption as queued because the post-request RLC falls to 39.9999%. RWA-PoB authenticates the attribution and integrity of institutional claims but does not independently prove the existence, ownership, or condition of off-chain assets. The prototype, test suite, datasets, and replication scripts are available at https://github.com/rischanlab/PoB.

Large language models (LLMs) frequently exhibit sycophancy: they adapt their answers to a user's stated beliefs or preferences instead of reporting what they hold to be true, which lowers factual accuracy and can amplify misinformation. This paper proposes a methodology for mitigating sycophancy that employs the Bayesian Truth Serum (BTS), a peer-prediction mechanism, as the reward in Group Relative Policy Optimization (GRPO) to fine-tune an LLM. BTS pays an answer for being surprisingly common, that is, more frequent among respondents than those respondents themselves predicted. We treat a group of responses from a model for one question as those respondents, so the reward is a function of the model's own outputs and fine-tuning needs neither labels nor preference annotations. We prove that in the large-group limit a sycophantic response earns strictly lower expected reward than an honest one. We also prove that if the entire group agrees in advance on a symmetric answering rule, it cannot earn a higher information score than under truthful reporting. On our true/false benchmark the reference model's answer-flip rate under user pressure decreases from 23% to 4%, and its accuracy under that pressure increases from 80% to 93%. Our reward outperforms SMART and is comparable to synthetic-data fine-tuning and to pinpoint tuning, all three of which train on labels. It spends considerably more compute in exchange, which makes it suitable when labeled data is scarce. Peer Truth Serum, which also pays a premium for a rare answer but elicits no prediction report, reproduces the effect. A peer-prediction reward computed inside a single GRPO group therefore reduces sycophancy without labels, and comparing mechanisms suggests that the premium paid for a rarer answer drives the effect.

This paper introduces and analyzes a novel algorithm—Truncated Fourier Filtering (TFF)—for the fast, high-order accurate evaluation of standard integrals and convolutions involving piecewise-smooth (possibly discontinuous) integrands over general $m$-dimensional domains ($m \ge 1$) employing an $m$-dimensional Cartesian grid. For an $N$-point discretization, the method runs at a computational cost of $\mathcal{O}(N)$ operations for standard integrals and $\mathcal{O}(N \log N)$ operations for convolutions, following, in either case, a one-time $\mathcal{O}(N \log N)$ precomputation step (not required in dimension $m = 1$). The core idea underlying TFF is to approximate the characteristic function of the integration domain by a truncated Fourier expansion of it over a suitably extended periodic domain, and to evaluate the resulting integrals via trapezoidal quadrature on a Cartesian grid with an appropriately chosen discretization size. Despite its conceptual simplicity, TFF attains high-order accuracy even for complex, possibly non-smooth or even non-Lipschitz geometries. A complete theoretical analysis is provided that establishes the superalgebraic convergence (i.e., convergence faster than any negative power of $N$) of the overall approach.

Mixture-of-experts (MoE) models combine specialized predictors under learned routing, offering a principled mechanism for leveraging heterogeneity in medical data. We present a hierarchical multimodal MoE for interstitial lung disease (ILD) classification that integrates a frozen, pre-trained imaging expert with structured electronic health records (EHR) via two-stage gating. A modality-level gate assigns patient-specific weights to imaging and EHR predictions, while a sub-gating module decomposes the EHR branch into clinically defined feature groups with learned, group-specific contributions. This design preserves stable imaging representations while enabling input-dependent clinical weighting and explicit EHR specialization. Under strict patient-level cross-validation, the model achieved the highest mean AUC among the evaluated methods (0.8750 +- 0.0443), compared with 0.8646 for imaging-only REN and 0.7685 for SwinUNETR. The framework extends interpretability across anatomical regions, imaging–EHR utilization, and clinically defined EHR feature groups.

Large-scale networked systems are increasingly vulnerable to adversaries that can infer system dynamics from a small set of compromised nodes. This paper addresses the problem of blocking such inference using limited state information. While existing state-feedback methods achieve observability blocking with eigenvalue preservation, they require full state access and are impractical for large networks. We propose multiple control strategies that operate under partial state measurements. The first approach employs output feedback to achieve observability blocking while preserving a subset of open-loop eigenvalues. The second approach leverages an observer to reconstruct the system state and enables full-state feedback control, preserving all eigenvalues and providing greater flexibility in sensor placement. We further extend the observer-based design to a distributed framework. Numerical examples demonstrate the scope and validity of the proposed methods.

Conventional discrete choice and machine learning models are estimated primarily from observational data and typically treat explanatory covariates as parallel inputs, providing no internal mechanism for determining how related attributes should adjust when one is deliberately changed. This paper proposes Neural-Bayesian Structure Learning (Neural-BSL), a framework coupling differentiable structure learning with random-utility-based discrete choice estimation in a single differentiable procedure. To prevent mutually exclusive choice outcome from distorting the recovered attribute structure, the observed choice is maintained outside the graph as an alternative-specific utility comparison, while the attribute structure and random-utility parameters are learned jointly. The learned structure enters the choice model through structure-weighted attribute interactions and provides the structural basis for propagating interventions through downstream attributes. An intervention is evaluated by updating the intervened attribute, propagating its model-implied downstream changes in topological order, and then recomputing utilities and choice probabilities. This yields both predicted mode-share responses and the associated changes in downstream traveler or trip attributes. We evaluate Neural-BSL using stated-preference data from Seoul and the revealed-preference data from London. Neural-BSL achieves predictive performance comparable to conventional benchmarks while recovering behaviorally coherent dependency structures. Across policy scenarios, propagating interventions through the learned structure changes the predicted redistribution across modes while exposing the downstream traveler and trip adjustments underlying those responses.

This study analyzes dynamic-key Learning-with-Errors (LWE)-based encrypted state-feedback control systems with time-varying encoders and decoders. Using a Lyapunov-based approach, we derive conditions on the time-varying encoder and decoder parameters that ensure both asymptotic stability and numerical safety by preventing overflow. The validity of the derived conditions is confirmed through numerical examples.

We introduce the wavemap: a spatial portrait of noise effects that assigns each site a per-noise-level arrival delay l_\gamma (v) and cross-entropy loss L_\gamma(v). These observables are exact at the lightcone frontier, where bond dimension \chi is small and the simulation is most faithful. Eigenvalue analysis of the composed gate-plus-noise Pauli transfer matrices confirms that the studied noise is pure amplitude damping: the spatial propagation pattern is entirely determined by the gate, making the wavemap a model-free noise diagnostic. We apply the multi-product formula (MPF) to recover the noiseless Pauli weight field from the noisy samples, subject to the Lieb-Robinson causal constraint nMPF <= nnl . Fitting time-adaptive coefficients \alpha(t) over the frontier recovers up to 55% of the information loss relative to the best noisy sample, exploiting the fact that the frontier is where truncation error is smallest. On an IBM heavy-hex lattice with heterogeneous hardware noise the method identifies an information-starved regime, pointing to calibrated synthetic noise as the next required experiment.

Medical vision-language models (VLMs) can appear reliable in-domain while failing when acquisition domain, paired supervision, or evaluation protocol changes. We study this failure mode as a representation-level blind spot relevant to epistemic intelligence, without claiming a formal estimator of epistemic uncertainty. Using NIH ChestXray14 and CheXpert, we first isolate source-only cross-dataset visual transfer from unsupervised domain-adaptation diagnostics. Using PadChest and OpenI, we then evaluate multimodal alignment under strict pair-index retrieval and quantify metadata-derived source-proxy information retained in frozen embeddings. Self-supervised visual initialization improves NIH-to-CheXpert transfer over supervised ImageNet initialization in matched ResNet-18 comparisons, whereas adversarial adaptation is useful only in a narrow regime and becomes unstable as adversarial pressure increases. Multimodal exact-pair retrieval remains low under external OpenI stress testing, and source-proxy information remains recoverable from learned representations. Qualitative nearest-neighbor and Grad-CAM analyses show clinically plausible cross-dataset structure and thoracic attention patterns in many cases, while device-heavy and false-positive cases remain ambiguous. Auxiliary architecture checks are task-dependent and do not support a universal backbone ranking. Overall, the study shows that apparent competence under a single protocol can conceal transfer, alignment, and shortcut-related failure modes, motivating stress-tested evaluation of medical VLMs under distribution shift.

The large-scale retirement of synchronous generators requires additional capabilities from inverter-based resources (IBRs) to ensure the stability and reliability of power grids. With the heterogeneous controls of IBRs, it is especially important to understand their behavior on the grid. This work proposes a simple data-enabled dynamic model to capture the small-signal dynamics of IBRs and formulate specifications for grid-forming (GFM) IBRs. The dynamic droop model is complementary to well-studied impedance models and extends the common definition of steady-state droop coefficients to dynamic droop coefficients that fully characterize the IBR small-signal response below the nominal line frequency (e.g., subsynchronous oscillations). We propose bounds on the gain and phase of the dynamic droop coefficients to encode minimum requirements for GFM IBRs to promote interoperability and minimize adverse interactions. The resulting specifications also provide some insights into the much-debated question of how to certify an IBR as GFM. Moreover, we also provide dynamic droop specifications for frequency control ancillary services that, e.g., clarify and generalize the notion of an IBR inertia response. Finally, common grid-following (GFL) and GFM controls as well as original equipment manufacturer (OEM) models are used to illustrate the results and showcase the use of dynamic droop coefficients as a tool to screen IBR dynamics for potential adverse interactions.

Industrial production modeling provides operational constraints for industrial users participating in demand response (DR) programs. Conventional modeling of the production process introduces binary variables to model the discrete operating points of industrial equipment, which can be computationally infeasible in large-scale DR applications. To reasonably model industrial users' operational constraints while balancing computational complexity and modeling accuracy, we developed a linear model of the industrial production process for evaluating DR applications. Numerical results verify the accuracy of the proposed model and its great improvement in computational efficiency over competing approaches.

System operators will probably allow virtual power plants (VPPs) to submit their feasible region (FR) for market clearing and dispatch. A VPP needs to determine its FR to submit as a whole based on the individual operation model of its internal distributed energy resources (DERs), which is an FR aggregation problem. Existing FR aggregation approaches rely on analytical methods, which have issues with generality and adaptability. In this paper, we propose a data-driven approach to approximate the energy-regulation FR of VPPs. It adopts the virtual battery model to approximate the aggregate FR of a VPP and determines the model parameters through inverse optimization based on generated multi-scenario operation data using the original operation model. Numerical tests verified the accuracy of the proposed method. We believe that our work helps to better leverage the flexibility of DERs.

Virtual power plants (VPPs) can aggregate distributed energy resources (DERs) to provide ancillary services for power systems, creating new profit opportunities. Ancillary services such as secondary frequency regulation require providers to have sufficient response capability to follow rapidly changing control commands. If overlooking the response requirement, the VPP will not be able to accurately measure its regulation capability, reducing its earnings in performance-based markets or risking disqualification. This paper integrates the requirement for fast-response capability into the operational framework of VPPs providing ancillary services. We leverage historical control commands to formulate chance constraints in the bidding model, mandating that the VPP's fast-response capability meets the requirement of ancillary services with a specified probability. Case studies verify that considering fast-response capabilities can enhance VPP operation.

Badminton is a compact but challenging domain for game AI: a player must choose a physically feasible shuttle trajectory, anticipate the opponent's interception, and recover to a court position whose value depends on the opponent's next response. The central challenge is that shot selection and recovery are not separable: the best recovery depends on the shot-induced opponent response, while the value of the shot depends on whether the hitter can cover the reply. This paper presents ShuttleArena, a physics-based singles badminton self-play environment that couples continuous shuttle flight, player interception, structured shot generation, and post-shot recovery. The policy uses role-conditioned outputs: a masked interception choice on receiver turns and a factorized hitter action over shot azimuth, shot elevation, shot speed, and recovery target, enabling interpretable tactical probes. Episodes are single rallies rather than full scored games, and training uses Proximal Policy Optimization (PPO) self-play against a staged checkpoint opponent pool with sparse terminal rally-outcome rewards and a factor-specific recovery update. Evaluation with frozen checkpoint play, controlled tactical probes, recovery ablations, qualitative rollouts, and a human-data sanity check shows competitive improvement together with interpretable opponent-conditioned changes in shot geometry and recovery behavior. The learned policies produce recognizable badminton-like structure while also reflecting the abstractions of the simulator, and the recovery intervention shows that learned recovery behavior is competitively important. These results suggest that physics-based racket sports are a useful testbed for interactive digital entertainment AI because they require agents to coordinate execution, positioning, and opponent-relative tactical value.

LLM-generated search queries are widely used to augment IR evaluation, yet they may contain concepts that presuppose answer-side document knowledge, violating the information-access boundary of pre-search users. Existing validation metrics, including overlap, diversity, and effectiveness, cannot distinguish rare human-tail variation from candidate answer-side intrusion. We introduce concept provenance, a framework that assigns query concepts to backstory-supported, human-central, human-tail, and candidate answer-side zones, operationalizing a boundary that retrieval metrics alone cannot detect. Applying concept provenance to 77,004 queries across 100 UQV100 topics, 8 LLMs, and 5 prompt conditions with two extraction pipelines, we obtain a cross-pipeline token-HCIR Spearman rho of 1.0 over five condition means. Candidate answer-side concepts constitute 7.40 percent of non-generic concepts and appear in 97 of 100 topics, with topic explaining approximately 67 percent of variance. Human validation yields 68.2 percent relaxed precision, revealing two mechanisms: knowledge intrusion at 45.5 percent and deployment intrusion at 45.0 percent. Diagnostic probes show disproportionate localized retrieval effects, with deletion effect size d = -0.47 compared with d = -0.34 for random deletion, but these concepts explain less than 2 percent of aggregate evaluation variance. Concept provenance therefore serves as a boundary-compliance diagnostic rather than an evaluation-shift predictor. Under the tested conditions, no prompt condition eliminates intrusion; post-generation concept-provenance selection achieves 99 percent elimination.

Recent open text-audio contrastive models (CLAPs) are typically trained with LLM-generated captions derived from tag datasets or web search results, which tend to be accurate but expressively narrow. As a complementary source, we explore human-written album reviews, specifically expert reviews from AllMusic: they exist at scale and carry narrative cues, evaluative adjectives, and scene framing that other sources lack. Since raw reviews are too noisy for direct use as captions, we first build a caption corpus with 24,5346 samples via an LLM preprocessing pipeline that identifies descriptive musical quotes and rewrites them into training-ready captions. We find that album review supervision yields the largest retrieval gains on a human-written caption benchmark (Song Describer), particularly for complex queries that other existing caption datasets leave uncovered. In addition, we revisit the training recipe and show that SigReg regularization, which encourages an isotropic Gaussian distribution in the embedding space, improves MLP probing across classification tasks, as well as text-to-music retrieval. The resulting model outperforms open CLAP-style baselines on text-to-music retrieval, zero-shot classification, and most MLP probing tasks. We release the review-derived caption dataset and model weights to support future research.

Continual knowledge injection is essential for keeping large language models up-to-date in a fast-evolving world. Existing methods rely on supervised fine-tuning (SFT), which memorizes injected facts in their training format but fails to generalize across paraphrasing, document combinations, and reasoning. To address this, we propose Golden-GRPO Injection (GRIN), a three-stage self-learning framework for continual knowledge injection. Golden-GRPO is a mixed-policy reinforcement learning algorithm designed specifically for knowledge injection, which injects a golden answer to provide learning signal even when on-policy rollouts fail on novel facts. We further introduce Blank and Counter, two document-level benchmarks targeting novel acquisition and counterfactual overwrite respectively, each evaluating single-fact recall, multi-source retrieval, and inferential reasoning. Our experiments establish a clear empirical claim: mixed-policy reinforcement learning enables knowledge absorption beyond what supervised fine-tuning can achieve. GRIN substantially outperforms SFT and mixed-policy RL baselines on the harder question types while matching them on basic fact recall.

Coding agents increase development velocity but also technical debt. Prior work reports only average effects across adopters, hiding wide differences between teams. We introduce RAMP (Repository AI Maturity Profile), a four-level cumulative maturity model grounded in version-controlled artifacts that teams commit to configure AI tools. RAMP runs from behavioral rules and coding standards through named agent definitions to multi-agent orchestration, with observed practice concentrated in the first three levels. Across 441 repositories the levels behave as a cumulative scale, and independent human annotation reproduces RAMP's repository-level labels on 97% of a held-out sample. Adoption is cumulative, forward-only, and set-and-forget: 73.8% of artifacts are committed once and never modified. Re-estimating an existing agent-adoption panel within each stratum, agents accelerate development regardless of maturity (28-38% more commits), but quality diverges: among agent-first repositories, where the contrast is identified, those without committed AI configuration show roughly twice the increase in cognitive complexity (+53% versus +27%) and 1.7x the increase in static-analysis warnings. Because maturity is observational, correlated engineering discipline or model capability may explain part of the gap; we present these findings as hypothesis-generating and release RAMP as a reusable instrument.

Understanding engagement in human interaction is essential for modelling assistive and context-aware AI systems. Previously, the MultiMediate challenge has focused on estimating momentary continuous engagement values in adult conversation across cultures, achieving high levels of generalisability across domains. However, engagement can be measured in different metrics and manifest differently between children and adults. MultiMediate'26 therefore expands the evaluated domain set to include the PInSoRo dataset of Child-Child and Child-Robot play sessions in English. These sessions were annotated categorically with task and social engagement, diversifying engagement types, age ranges and interaction types. In this paper, we describe the MultiMediate'26 challenge and its goals, introduce the NoXi, NoXi-J, MPIIGI, and PInSoRo datasets, showcase extracted multi-modal features, present our baseline model for engagement estimation across all domains, and discuss challenge submissions and approaches. In an effort to provide a transparent and welcoming challenge environment, baseline code, leaderboard, datasets, and extracted features are publicly available at https://github.com/hcmlab/MultiMediate_26.

Though LLM-simulated practice may support psychotherapy education, open questions remain about which system features provide the most educational value. We developed an AI-simulated practice tool for use in psychotherapy classrooms that provides speaking practice with LLM-based patients, followed by a post-practice activity that includes feedback from a fine-tuned LLM and written reflection exercises. We deployed the tool in a graduate psychotherapy course (n=25, 5 weeks) using a micro-randomized trial (MRT) - a method which can estimate the causal excursion effect of an intervention by leveraging longitudinal repeated randomization with participants. Testing four conditions crossing AI feedback (present/absent) with reflection granularity (utterance/session-level), we found surprising interaction effects on student engagement and educational value: utterance-level reflection paired with AI feedback significantly outperformed all conditions. Students valued comparing the AI suggestions to their own rather than passively accepting them. As the first MRT in a health education setting testing an LLM-simulation system, this work demonstrates that MRTs, a method mostly used in mobile intervention research, offers a viable evaluation paradigm for AI systems deployed in health education settings.