Papers

We study the optimization of convex objectives with $(L,\kappa-1)$-Hölder-continuous gradients in $\ell_q$ over $R B_p^d$, $1<\kappa\le 2$. (MG26) provides selectors with a movement bound for the problem of chasing high-dimensional convex nested sets for every $p<q$ and generally reduces Lipschitz convex optimization to bounds on the movement of selectors. We couple that movement with Hölder descent yielding a polynomial-runtime first-order method whose feasible output, in the high-dimensional regime $T\le d$ and for $p<\min\{q,2\}$, has error $$ \widetilde O_{\kappa,p,q}\!\left( \frac{LR^\kappa}{T^{\kappa(1+1/p-(1/q-1/2)_+)-1}} \right), $$ after $T$ queries to a first-order oracle, solving the COLT 2015 open problem of (Guz15), up to logarithmic factors. At $(p,q)=(1,2)$, the rate is $\widetilde{O}(LR^\kappa/T^{2\kappa-1})$, including $\widetilde{O}(LR^2/T^{3})$ cubic decay in the smooth case.

Vision-Language Models (VLMs) have recently been proposed as promising tools for face recognition, as they can produce natural language explanations alongside similarity scores. This capability is considered appealing for face comparisons in forensic contexts, which require decisions to be transparent and auditable. However, existing evaluations of VLMs for that use case focus mostly on recognition accuracy, while the validity of generated explanations remains unquantified. In this work, we introduce a benchmarking framework for VLM-based face recognition that treats explanation quality as a core evaluation axis. We propose two criteria that explanations should satisfy: relevance, i.e., reliance on identity-stable facial features; and faithfulness, i.e., alignment with the visible image content without hallucinated features. We jointly develop a methodology enabling the quantification of relevance and faithfulness of evaluated models, based on constraining model outputs to a structured explanation format that supports automated querying and auditing. Using this framework, we benchmark several families of open-weight VLMs, jointly evaluating face verification accuracy and explanation quality. Our results highlight remaining shortcomings of produced explanations, and emphasize the need for such explanation quality metrics to get a complete picture of model performance. The proposed benchmark and open-source evaluation harness provide a foundation for proper benchmarking and future fine-tuning of explainable face recognition systems.

As an alternative to the additive and extremal biases of average and max pooling, we introduce Geometric Mean Pooling (GMP), a signed pooling operator that combines the product of feature signs with the geometric mean of feature magnitudes. Motivated by local-to-global composition in quantum many-body physics, GMP retains both joint sign information and a characteristic multiplicative scale without introducing learnable pooling parameters. We show that non-overlapping hierarchical GMP preserves the corresponding global multiplicative statistic and evaluate it on synthetic sequence tasks, iterative coarse-graining, image classification, and molecular lipophilicity regression. On the synthetic tasks, GMP recovers product-based signals more accurately than average and max pooling and maintains predictive performance under the tested levels of multiplicative input noise. On image and molecular data, however, its effectiveness depends on the representation, target parameterization, and placement of local and global pooling. These results position GMP as a complementary, regime-dependent inductive bias for tasks in which equal-weight multiplicative composition is plausible, rather than as a universal replacement for standard pooling operators.

Periodic inspection of deployed wireless devices is necessary because unauthorized hardware replacement may preserve communication functions, credentials, and logical identity, making anomalous devices difficult to detect. Such inspections are conducted under controlled measurement conditions to verify that each device remains consistent with its enrolled hardware state. Conventional radio-frequency fingerprint (RFF) may provide insufficient separation when replacement hardware closely resembles legitimate hardware, while a polarization fingerprint (PF) constructed at one observation direction may miss spatially nonuniform polarization changes. This paper proposes the spatio-frequency polarization fingerprint (SFPF), which jointly represents complex polarization responses over multiple frequencies and observation directions; conventional PF is its fixed-direction slice. We derive SFPF formation from hardware-dependent modal excitation, directional far-field radiation, and polarization projection. A first-order sensitivity analysis shows that the response to the same hardware change varies with both frequency and direction, motivating joint spatio-frequency acquisition. Electromagnetic simulations confirm the nonuniform spatio-frequency sensitivity and show that, under the same observation budget, SFPF improves normalized distance, Fisher score, and the inter-/intra-class ratio over PF by 17.7%, 45.8%, and 11.3%, respectively. Experiments show that SFPF consistently outperforms RFF and PF over 0–20~dB. At 15–20~dB, SFPF achieves anomalous-device F1 scores of 87.3–90.4% and AUROC values of 85.4–95.5%.

We introduce Chronosphere, a spatio-temporal neural field that learns representations of climate. A central challenge in geographic representation learning is modeling environmental processes whose spatial and temporal complexity varies widely. Yet existing location encoders typically fix a single level of detail everywhere. Global bases such as spherical harmonics spread capacity uniformly across space and time. Localized bases resolve only predefined regions. Learned tessellations adapt, but are inefficient at representing higher frequencies. Chronosphere unifies these approaches, pairing an adaptive tessellation of learnable sites on the spacetime torus $S^2\times S^1$ with a shared bank of local basis functions. Both where capacity is placed and how much detail each region carries adapt to the data, across space and time. Trained to reconstruct climatology, Chronosphere matches or leads state-of-the-art location encoders across spatial and temporal tasks, with the largest gains under spatial and temporal transfer.

We propose a filtered time stepping finite element scheme for curve shortening flow of open and closed curves in arbitrary codimension that is second-order accurate in time. Open curves are assumed to evolve inside a given domain $\Omega \subset \mathbb R^n$, $n\geq2$, and meet the external boundary $\partial\Omega$ orthogonally. We prove optimal error bounds for the $L^2$– and $H^1$–norms. In practice only a single linear system needs to be solved at each time step. Numerical experiments confirm the accuracy and practicality of the introduced method, including an asymptotic equidistribution property.

Modern deep learning models achieve impressive generalization through over-parameterization, but this paradigm often struggles with overfitting and memorization in few-shot regimes. Neural Cellular Automata (NCAs) offer a highly parameter-efficient alternative, yet research has focused primarily on their output, leaving the role of their internal hidden channels largely unexplored. In this paper, we investigate the internal dynamics of NCA hidden channels and introduce a novel transfer-learning mechanism that injects a pretrained teacher's hidden states into a student model to guide early optimization. Evaluated on few-shot and scale-variant MNIST benchmarks, NCAs outperform comparable recurrent and feed-forward architectures, demonstrating superior generalization with a minimal parameter budget (~9,800 parameters). Mechanistic analysis reveals that the hidden channels decouple feature extraction from uniform classification consensus by absorbing morphological complexity and converging to mutually orthogonal states. Furthermore, we demonstrate that these hidden channels capture general, scale-invariant topological primitives rather than class-specific templates. This allows a student model to achieve strong few-shot performance on unseen classes using features transferred from a teacher trained only on a subset of digits (0-5). Our results highlight the potential of utilizing hidden-state dynamics as a robust, decentralized computational substrate for parameter-efficient transfer learning

Hand tracking offers natural VR interaction but lacks controllers' inherent tactile feedback, while haptic feedback across input methods remains understudied. We conducted a participant study comparing vibration, impulse, and no feedback in hand- and controller-based VR interaction across grasp-, pinch-, and tap-like gestures assessing performance, workload, and preference. A custom glove and controller-mounted device provided both feedback types, aiming for consistency across input modalities. Controllers were faster for grasp and pinch, whereas hand tracking was more accurate for pinch and preferred overall. Haptics had limited performance effects, although impulse reduced grasp accuracy with controllers. Participants preferred vibration and impulse over no haptic feedback, favouring vibration overall. Our findings reveal a more nuanced relationship between haptic feedback, input device, and interaction context than suggested by previous work.

We prove fully discrete convergence of the classical Barrett–Garcke–Nürnberg (BGN) scheme for curve-shortening flow of smooth embedded closed planar curves. The main obstruction is that the mass form controls only normal motion, whereas the tangential motion is determined implicitly by the curvature equation and governs the parametrization. The usual length-decay estimate therefore does not control perturbations of the full position update. We separate temporal and spatial errors through the time-semidiscrete BGN solution. A scalar normal resolvent and exact curvature and length identities yield uniform regularity and first-order time convergence. For the spatial analysis, an adapted normal–tangential norm gives a near-contractive estimate for the linearized update. Gauss–Lobatto cancellations and an exact covariance identity for the assembled nodal normals produce an $H^1$ one-step defect of order $h^{k+1}$, while an exact difference identity controls the nonlinear remainder. For every fixed $k\ge1$, including the original piecewise linear method, we obtain the matched $W^{1,\infty}$ error bound $C(\tau+h^k)$ on periodic quasi-uniform meshes with $h\le c\tau^2$. All discrete steps are uniquely solvable, and the numerical curves remain regular and embedded. To the best of our knowledge, this is the first fully discrete convergence result for the classical BGN curve-shortening scheme without additional stabilization.

Wafer map defect recognition is commonly formulated as a fixed-taxonomy classification problem that assigns each wafer to a single defect class. However, some wafers exhibit morphologies near class boundaries, for which forcing a single prediction may be less informative than providing plausible diagnostic alternatives. This paper proposes a morphology-aware ambiguity learning framework that supports three diagnostic actions: automatic single-class diagnosis, assisted diagnosis with two plausible defect classes, and full review. Using the radial, angular, and geometric characteristics of training wafer maps, the framework constructs a class-level ambiguity matrix representing defect-class pairs with similar morphology and plausible diagnostic alternatives. It guides the model to learn plausible alternative classes rather than treating all incorrect classes equally. During inference, the matrix determines whether an uncertain prediction can be represented by a meaningful two-class diagnostic set or should be escalated for full review. Experiments on WM-811K show that the proposed framework outperforms conventional approaches in defect recognition and diagnostic decision support, providing meaningful two-class alternatives while reserving full review for cases with unresolved ambiguity. Illustrative cost analyses further show the potential cost advantage of the proposed routing strategy. The diagnostic behavior of the framework remains consistent across different backbone architectures.

Empirical evaluation is central to recommender-systems (RecSys) research, but turning experimental designs into executable code remains a manual and error-prone task. We present AutoRecLab, a Python-based autonomous RecSys lab that automates RecSys experiments from natural-language prompts. Given a research idea, AutoRecLab derives explicit experiment requirements, builds and validates a prototype, and iteratively expands it into the requested full experiment. The workflow combines retrieval-augmented generation (RAG) for documentation lookup, static type verification, and execution-steered tree search. In our demonstration, AutoRecLab autonomously implements an explicit-to-implicit feedback conversion study. In a baseline comparison across six algorithms and three datasets, 8 of 9 runs succeed at an average cost of approx- imately $1 per run with GPT-5.4-mini.

The transmission eigenvalue problem plays an increasingly important role in inverse scattering theory. Although significant progress has been made in developing efficient numerical methods for the problem over the last two decades, its numerical treatment for curved domains in $\mathbb{R}^d$ ($d=2,3$) remains challenging. In this paper, we introduce and analyze a geometrically flexible and $H^2$-conforming isogeometric method for solving a fourth-order, quadratic and non-self-adjoint eigenvalue problem arising from the Helmholtz transmission eigenvalue problem. Using the spectral approximation theory for compact non-self-adjoint operators, we derive optimal error estimates for the discrete eigenvalues and eigenfunctions. Numerical results for the two- and three-dimensional benchmark problems, including those defined in curved domains, are presented to verify our theoretical results, and to demonstrate the advantages of the method over existing numerical methods in terms of both accuracy and geometric flexibility.

Nearly 90% of drugs entering clinical development ultimately fail, despite billions of dollars in investment. Pharmaceutical companies therefore rely on clinical development planning (CDP) and probability of technical and regulatory success assessment to anticipate development risks, yet these decisions remain labor-intensive and subjective, requiring experts across clinical science, statistics, regulatory affairs, and competitive intelligence to jointly acquire, synthesize, and reason over heterogeneous evidence. Here, we introduce TrialAtlas, a memory-augmented multi-agent research organization for CDP that mirrors this collaborative process by coordinating specialized agents for literature synthesis, competitive trial intelligence, regulatory precedent analysis, and integrated reasoning over trial design and development risk. TrialAtlas further learns from historical clinical trials and regulatory outcomes, including prior New Drug Applications (NDAs), to ground its decisions in accumulated development experience. To evaluate these capabilities in an authentic regulatory setting, we introduce TrialAtlasBench, constructed from 291 FDA Complete Response Letters and spanning three practical tasks: detecting trial design deficiencies, recommending actionable design improvements, and predicting technical and regulatory success. TrialAtlas achieves an F1 score of 50.0% for deficiency detection, outperforming the strongest baseline by 6.1 points, and reaches 85.3% balanced accuracy and 84.7% F1 for prediction of technical and regulatory success, improving over the best baselines by 6.7 points in balanced accuracy and 12.0 points in Cohen's kappa. In expert evaluation, 86.4% of TrialAtlas-generated concerns were judged valid, compared with 83.1% for OpenAI DeepResearch and 59.3% for Gemini DeepResearch.

Large language models (LLMs) have achieved state-of-the-art performance across a wide range of tasks, motivating two important aspects of deployment: inference efficiency and output provenance, which can be tackled by speculative sampling and watermarking, respectively. However, recent works have shown that combining these two goals is highly nontrivial and can be potentially impossible. In this work, we develop a novel multi-draft speculative sampling algorithm based on Poisson processes that improves the frontier of this fundamental trade-off. The proposed algorithm has strong sampling efficiency on its own and, more interestingly, is naturally watermarkable: we can embed an unbiased watermark without degrading speculative acceptance. Moreover, our algorithm is based on an exact list-coupling-without-communication scheme, which yields a drafter invariance property that benefits both sampling and watermarking. It is the first multi-draft, drafter-invariant speculative sampling scheme that maintains both watermark strength and sampling efficiency, and we experimentally verify its strong performance in both aspects.

LLMs are increasingly used in social simulations for socially interactive agents and robots, offering more flexibility than rule-based systems. However, even though they mimic human behaviour very well, there is a persistent alienness to them. This work investigates whether personality-aware fine-tuning can reduce this gap by improving the consistency and controllability of personality-conditioned dialogue generation compared with instruction prompting alone. We fine-tune two small open-weight LLMs, Qwen2.5-7B-Instruct and Ministral-8B-Instruct, using a corpus that combines personality-labelled social media posts and dialogues to create a personality-based dialogue engine for social simulation. The resulting models are evaluated across multiple social interaction scenarios using three independent LLM judges, which assess personality fidelity and provide evidence-based behavioral interpretations. We additionally quantify inter-rater agreement and lexical characteristics of the generated dialogue. Results indicate that fine-tuned models are not better at role-playing different personalities than their respective baseline models. However, low inter-rater agreement limits the confidence with which these results can be interpreted. Concerning the quality of generated texts, fine-tuned models are mostly comparable to the baselines, with fine-tuning improving the linguistic diversity of the Qwen models. While the results appear generally usable and the baseline models offer the best overall performance, future studies should place greater emphasis on the quality and domain alignment of training data for accurate personality role-playing.

Secondary control and the State-of-Charge (SoC) balance control are important control objectives for battery energy storage systems (BESSs). In this brief, a communication weight allocation method based on the centrality of eigenvectors is designed for a connected and directed graph, which results in the adjacency matrix having a given eigenvector. Subsequently, a distributed secondary voltage controller and an SoC balancing controller are designed for droop-controlled BESSs to achieve voltage leader-following multiconsensus and SoC balancing, respectively. It is worth mentioning that under the designed voltage secondary control scheme, only a single leader is needed to achieve voltage multiconsensus control. In addition, the capacity information/droop coefficient does not need to be transmitted in the communication network to achieve power sharing according to capacity and SoC balance. For SoC balance control, the control gain is also well analyzed to ensure stability. The relevant simulations verify the effectiveness of the designed scheme.

Gratitude technologies support well-being by prompting reflection on what people appreciate. But gratitude does not serve the same purpose in every circumstance: as life situations change, so does what people seek from it, and whether it feels appropriate at all. To understand how technology can adapt to and support such shifts, we conducted retrospective, artifact-elicitation interviews with 17 adults who had practiced gratitude for one to fifteen years. Participants' appraisals of their situations shaped what they needed, yielding six recurring practice patterns, including a boundary where gratitude felt forced. We contribute the Adaptive Gratitude Practice Model, which explains how appraisals shifted even within the same life situation, how participants adapted activities, modalities, and rhythms, lapsed under competing demands or emotional unreadiness, and resumed when gratitude again felt useful. Additionally, we derive design implications for situated support, self-understanding through past records, and relational care with changing life situations.

On-device inference is booming, but the momentum is almost all in language models. Diffusion pipelines are memory hungry, latency-sensitive, and require orchestrating an embedder, a transformer, a decoder, and often further postprocessing that is not as standardized as LLM inference loops are. We navigate the trade-off between performance, quality, and model footprint to reach as many client devices in the wild as possible. We make three contributions: an embedding translator that maps a small text encoder into a large encoder space to cut weight and latency; a reproducible sweep recipe for navigating the speed/quality/memory triangle in diffusion pipelines; and an interactive on-device image generation editor achieving sub-second TTFI on recent GPUs.

The balanced butterfly is a fundamental primitive for analyzing signed bipartite graphs and provides a basis for studying higher-order structural properties, such as clustering coefficients and community structure. Despite its importance, existing approaches primarily rely on serial algorithms for balanced butterfly counting, which become inefficient on large-scale graphs. To address this limitation, we propose a distributed algorithm, D-BBC, based on a hybrid MPI+TBB framework that exploits MPI for inter-process communication and Intel TBB for intra-node parallelism. We conduct an experimental assessment of the proposed approach across 15 real-world datasets. Experimental results demonstrate that, on a single-node distributed system, D-BBC achieves average speedups of 1321x and 16.2x over the serial BB2K and multi-core M-BBC implementations, respectively. Furthermore, D-BBC achieves a maximum speedup of 23.58x over the distributed baseline S-Monarch in terms of end-to-end execution time. These results demonstrate the efficiency of the proposed distributed approach and its potential to enable high- performance signed motif analysis on large-scale bipartite graphs.

Long transcripts are costly inputs for downstream NLP systems and often contain irrelevant context. We study query-conditioned topic localization: predicting the sentence span in a transcript that best addresses a topic-title query. To improve span localization, we reuse ASR encoder states as sentence-level representations and fuse them with textual embeddings. This lets lightweight span locators exploit speech information without running a separate audio encoder. Experiments on two public datasets show consistent gains over text-only baselines, especially under strict boundary-matching criteria. Cross-dataset experiments further indicate that the benefits are strongest for structured or semi-structured speech, while gains on spontaneous speech are limited and mixed.

This study examines tool-augmented Large Language Model (LLM) systems for supporting Root Cause Analysis (RCA) of nightly test failures at Westermo Network Technologies AB. Nightly test executions produce heterogeneous test data and logs that practitioners currently inspect manually across multiple sources. We implemented an RCA workflow in single-agent and orchestrated multi-agent configurations, both with access to test metadata and logs. An exploratory industrial case study used two real failure scenarios. Six practitioners evaluated the scenario reports through a survey and focus group, and operational measurements were collected from 120 repeated executions. The evaluation covered practitioner-perceived correctness, reasoning quality, fix realism, clarity, usefulness, and trust, as well as cost, duration, and consistency. Neither configuration showed a consistent practitioner-perceived quality advantage across the two scenarios. The single agent system generated reports faster and at lower cost, making it the more practical baseline in this context. The potential benefits of agent architectures require further evaluation in more complex scenarios.

Tree Decision Diagrams (TDDs) are a data structure recently introduced by Capelli et al. (SAT 2026). They are structured along a vtree and the size of their canonical form lies between Ordered Binary Decision Diagrams (OBDDs) and deterministic structured DNNF circuits (d-SDNNFs). While the succinctness gap between TDD and d-SDNNF is exponential, only a quasipolynomial separation between OBDD and TDD has been shown and it was left as open question whether this is optimal. We answer this question affirmatively by showing that every TDD can be transformed to an equivalent OBDD of quasipolynomial size. Although this might be seen as a weakness, our second result shows that TDDs share another desirable property with OBDDs that is not known to hold for d-SDNNF: Given a TDD and another target vtree, it is possible to construct the minimal and canonical TDD respecting the new vtree in time polynomial in the input and output. As a result we also obtain that the equivalence test between TDDs over different vtrees can be done in polynomial time.

Frontier language models now produce professional deliverables that expert graders judge to match human work on a substantial share of economically valuable tasks, yet most enterprise GenAI initiatives fail to show a measurable business effect and a large fraction of agentic projects are expected to be cancelled. We argue that this is substantially a measurement problem: public benchmarks answer "what can the model do?", whereas a deployment decision requires "is this workflow fit, reliable, safe and worth scaling - here, on our data, under our controls?". We present EnterpriseVal, a use-case-level evaluation system that closes this gap. It comprises (i) a formal specification of the use case and of the frozen socio-technical configuration under test, model, prompts, retrieval, tools, guardrails and human oversight, with an autonomy level and consequence tier that jointly set the required evaluation intensity; (ii) a metric catalogue spanning fidelity, utility, efficiency, reliability, assurance and oversight; (iii) a grading protocol that scales blinded expert judgement with calibrated LLM-as-judge scoring through prediction-powered inference; (iv) a two-tier threshold gate, stated as an executable algorithm, that maps metric vectors with confidence bounds to REJECT/CONDITIONAL/SCALE decisions; and (v) a value-and-risk model in which the reviewer catch rate is a measured parameter. We report a pilot across three workflows in a global bank. In credit-memo drafting, human-graded citation precision reached 88% and hallucination rate 1.6% for the best model against gates of 70% and 5%; in procedure transformation, analyst refinement effort fell from an estimated 27.4 to 2.9 hours per document. We separate established results, documented pilot evidence, the proposed system and open hypotheses, and specify the experiments required for full validation

We show that fixed-parameter tractability and polynomial-time solvability coincide for homomorphism problems specified by classes of pattern hypergraphs of bounded incidence degeneracy or bounded primal independence number, assuming $FPT \ne W[1]$. In both cases, tractability is characterised by bounded fractional hypertree width. This generalises a result of Grohe [JACM 2007] that established such an FPT-PTIME collapse for bounded-arity structures. We build on the fractional balanced-separator framework and rounding theorem of Korchemna, Lokshtanov, Saurabh, Surianarayanan, and Xue [FOCS 2024] to prove near-linear bounds on fractional hypertree width ($fhw$) in terms of adaptive width ($adw$). Namely, for every hypergraph $H$ $$fhw(H)=O(\mu(H) adw(H)\log adw(H)), \quad fhw(H)=O(\log \alpha(H) adw(H)\log adw(H))$$ where $\mu(H)$ is the degeneracy of the incidence graph of $H$ and $\alpha(H)$ the independence number of the primal graph of $H$. As a consequence, we obtain a corresponding FPT-PTIME collapse for exact homomorphism decision and counting on every class in which either incidence degeneracy or independence number is bounded. This includes a number of natural classes such as pattern hypergraphs of bounded degree. Additionally, we observe that for every r.e. class of pattern hypergraphs, fixed-parameter tractability of the parameterised homomorphism problem implies quasipolynomial-time solvability of the corresponding unparameterised problem, assuming $FPT \ne W[1]$. Our results rely on the characterisation of fixed-parameter tractable decision of homomorphism problems in terms of submodular width of Marx [JACM 2013], conditional on ETH. In light of the very recent potential refutation of the ETH, we show that the characterisation of Marx holds also under the assumption that $FPT \ne W[1]$.

Social-navigation algorithms are often evaluated under a fixed pedestrian-behavior distribution, despite substantial variation in pedestrian responses to robots across individuals and social contexts. We introduce PopNavShift, a matched simulation framework for stress-testing social-navigation strategies under pedestrian population shifts. PopNavShift constructs population-conditioned pedestrian motion profiles by prompting Gemini 3.7 Flash with 600 synthetic persona records from MatrAIx Persona 1M and deterministically mapping the responses into bounded motion parameters. It then compares three representative navigation strategies, reactive avoidance, early yielding, and reciprocal collision avoidance, across eight population conditions and 7,488 matched robot runs. In a matched intervention on the same 202 personas, changing only time pressure reverses 8.6% of controller rankings based on robot travel time, but 22.4% based on mean pedestrian delay and 23.9% based on worst-decile delay. Across population conditions, this sensitivity is greater for pedestrian burden than for robot travel time and increases in spatially constrained settings; the same qualitative pattern persists under a second pedestrian dynamics model. These findings support evaluating navigation strategies across behavioral populations using both robot performance and pedestrian burden.

Clustering high-dimensional data is a fundamental task in unsupervised machine learning with applications to a variety of domains. In the centralized data scenario, this task is commonly solved using deep clustering methods that utilize deep neural network architectures to learn clustering-friendly latent space representations. In Federated Learning, where data is distributed between clients and is private, deep clustering methods are less explored. In particular, recently introduced federated deep clustering methods, despite showing very promising performance, still fall short in reliably providing good performance if data across clients are non-identically-independently distributed. In this work, we introduce a generalization of Deep Clustering Networks to the federated scenario, named FedDCN, that simultaneously optimizes a reconstruction loss and a clustering loss. To ensure robustness and latent space alignment in non-identically-independently distributed data scenarios, FedDCN generates synthetic data augmentations, and its learning objective includes a geometric regularization for latent space alignment. Through experimental evaluation, the effectiveness of the approach under IID and non-IID assumptions is demonstrated, and future research directions are identified.

Blind and low-vision users often face challenges when locating and physically acquiring objects in unfamiliar indoor environments. Existing vision-language-model-based assistants can provide semantic descriptions but may introduce latency, hallucinations, and guidance that is poorly aligned with embodied action. We present Touvigation, a hands-free object acquisition system that combines vision-language understanding with persistent local spatial modeling to provide low-latency, body-relative guidance. Drawing on formative interviews with eight blind and low-vision participants, we design a multi-stage guidance framework that adapts spatial references as users transition from orienting, to walking, to reaching and tactile verification. We evaluated Touvigation with 12 blind and low-vision participants against a multimodal large-language-model assistant and unassisted search. Touvigation achieved 100% task success, compared with 58% for the multimodal assistant and 85% for unassisted search, while reducing completion time and cognitive workload. Our findings demonstrate how persistent spatial grounding and adaptive embodied guidance can improve object acquisition for blind and low-vision users.

Speculative decoding accelerates LLM inference by drafting multiple tokens in parallel, with tree-based methods further improving efficiency through hierarchical structures. Dynamic-tree methods such as EAGLE-3 perform well under greedy decoding via deterministic top-K expansion and global pruning. However, in stochastic decoding (T>0), this mechanism collapses the draft distribution into one-hot probabilities, causing a severe drop in acceptance rate. This creates a dilemma: dynamic-tree methods sacrifice stochastic sampling to preserve context-aware topology, while static-tree methods preserve stochastic sampling with context-agnostic structures. The issue arises because the same probability distribution is used for two conflicting tasks: constructing the tree and verifying tokens. This coupling makes direct injection of randomness challenging due to the resulting stochastic process. We resolve this by decoupling these roles: RheoSampling assigns a token sampled from the draft distribution a proxy probability for tree expansion and pruning alongside its true sampling probability for verification. Specifically, we inject a sampled token among the deterministic top-K slots and treat it with different probabilities during construction and verification, making RheoSampling the first dynamic-tree method with both context-aware top-K construction and stochastic sampling while maintaining losslessness. We establish the lossless guarantee through an equivalence-class analysis that compresses the stochastic tree space into tractable classes. An OT-based verification strategy and a sparse draft mechanism ensure that theoretical gains translate into practical efficiency. Experiments across LLMs and benchmarks demonstrate improvements in acceptance rate and speedup over state-of-the-art dynamic tree methods. This framework may provide a template for analyzing stochastic tree structures.

We initiate the study of $\alpha$-fair prophet inequalities. This interpolates between utilitarian welfare $(\alpha=0)$, Nash welfare $(\alpha=1)$, and Rawlsian max-min fairness $(\alpha\to\infty)$. Given the non-linearity of the objective, it matters when the expectation is applied. For instance, for the Rawlsian objective, it matters whether we aim to maximize $\min \mathbb{E}[u_i]$ or $\mathbb{E}[\min u_i]$. We refer to the former as the ex-ante model, and the latter as the ex-post model. For ex-ante fairness, full distributional knowledge yields a tight competitive ratio of exactly $1/2$ for every $\alpha\ge 0$. Under sample access, $O(n\log n)$ samples per distribution suffice for a constant competitive ratio when $\alpha\in(0,1]$. In contrast, for every $\alpha>1$, no finite number of samples improves upon the trivial $1/n$ guarantee. Thus, unlike in the utilitarian setting, full-information and sample-access prophet inequalities become fundamentally separated. For ex-post fairness, under full information, we obtain a uniform constant ratio for all $\alpha\in(0,1)$, while for every $\alpha>1$ the competitive ratio collapses to $1/n$. In the sample-access model, one sample per distribution suffices for each fixed $\alpha<1$, but no sample budget depending only on $n$ yields a uniform constant guarantee as $\alpha\to 1$. Beyond these phase transitions for $\alpha$-fairness, our results open the door to a broader theory of prophet inequalities for non-linear welfare objectives.

While object detection has advanced through improved architectures and open-vocabulary models, we provide strong evidence that benchmark quality is limited by annotation incompleteness. Across four widely used datasets (COCO, Pascal VOC, Cityscapes, KITTI), re-annotation reveals substantial increases in annotated objects (e.g., up to +60% on KITTI and +40% on COCO), driven primarily by previously unlabeled small, occluded, or densely packed instances. While some differences arise from dataset-specific annotation conventions, we consistently find that missing annotations are the main source of label errors across all datasets. To achieve high data quality, we introduce a scalable annotation pipeline that emphasizes high recall and captures ambiguity through soft labels aggregated from at least 11 annotators per object. The resulting annotations improve coverage and align well with human calibration. We show that benchmark performance is highly sensitive to annotation quality, although model rankings remain largely stable. We introduce two large-scale benchmarks: (i) an uncertainty-aware object detection benchmark, and (ii) a label error detection benchmark grounded in real label errors. We show that current detectors are strongly depended on annotation quality and are misaligned with human perception. Current label error detection methods, which have been shown to perform well on synthetic noise, struggle to achieve high recall and precision on real label errors. Our results highlight the need for future object detection benchmarks to move beyond deterministic annotations toward high-recall, uncertainty-aware evaluation that maximizes valid instances and better reflects real-world ambiguity.