Papers

Safe crowd navigation under distribution shift requires uncertainty representations that capture structured human-motion prediction errors and safety objectives that account for rare but consequential failures. Existing uncertainty-aware methods typically represent prediction errors using isotropic regions, which can be either overly conservative or poorly aligned with directional motion uncertainty. We introduce a risk-aware navigation framework that uses anisotropic conformal ellipsoids to translate structured prediction uncertainty into an episode-level conditional value-at-risk (CVaR) signal that regulates the Lagrangian safety penalty. In particular, adaptive ellipsoidal conformal prediction (AECP) captures directional prediction errors and adaptively calibrates uncertainty regions under distribution shift, while the resulting CVaR-regulated navigation policy is optimized using Lagrangian proximal policy optimization. We evaluate our proposed method under both in-distribution settings and out-of-distribution (OOD) settings involving shifts in pedestrian motion patterns. Compared with state-of-the-art baselines, our method maintains competitive in-distribution performance while improving success rates by 5.68-7.44 percentage points and reducing collision rates by 5.44-6.64 percentage points across OOD settings. We further deploy the trained policy without fine-tuning on a physical robot with onboard perception and CPU-only inference, showing that the full pipeline is feasible in physical crowd navigation.

Large language models show promise in clinical reasoning, but psychiatric interviewing requires guiding an evolving conversation. Their ability to carry out this interactive assessment remains less studied. We present PsyCIDRA, a dual-agent framework linking free-form psychiatric interviewing with diagnostic reasoning for expert review. Its interviewer agent uses tools to maintain working notes, load expert-written skills, and retrieve ICD-11 references to guide inquiry. Its diagnostic reasoning agent then receives the completed interview transcript and reports hypotheses alongside supporting, conflicting, and missing evidence, withholding a final hypothesis when none is sufficiently supported. Using patient profiles generated with PsyCPG, we first evaluate PsyCIDRA in simulation. Across four models on 53 evaluation cases, it achieves higher diagnostic agreement than direct prompting. On 81 held-out simulated cases, rank-1 accuracy is 60.5% versus 51.9%. In a blinded study of 101 human participants in separate arms, PsyCIDRA agrees with psychologists on whether to propose a diagnostic hypothesis in 79.6% of cases, compared with 65.4% for direct prompting. Together, these findings support the potential of LLM agents to assist psychiatric assessment through free-form dialogue. By examining diagnostic reasoning, interview quality, and safety together, this study contributes to understanding the capabilities and limitations of psychiatric interview agents.

Remotely operated vehicles (ROVs) are widely used to explore and inspect underwater environments such as caves, shipwrecks, and submerged infrastructure. These missions require accurate 3D understanding of the surrounding environment, which depends on both reliable vehicle localization and metric scene reconstruction. However, external positioning is often unavailable underwater, requiring small ROVs to rely primarily on onboard perception. Optic vision provides rich visual and geometric information but suffers from scale ambi- guity and trajectory drift, whereas 2D imaging sonar provides metric range but incomplete 3D geometry. Existing underwater reconstruction approaches typically address these limitations separately or assume known sensor poses, leaving localization and reconstruction disconnected. We present SURGE, a camera sonar framework that jointly estimates the ROV trajectory and target location by integrating visual and acoustic observations within a factor graph, then uses the recovered metric poses for sonar Gaussian splatting. Experiments on real underwater RGB sonar observations show that SURGE substantially improves localization consistency over conventional vision based pose estimation and produces a more compact, natively metric reconstruction than RGB Gaussian splatting baselines.

Combinatorial Thompson sampling (CTS) draws independent posterior samples for every arm, so its exploration dynamics ignore any relation among arms. We study a minimal change to those dynamics: an LLM is queried once for a partition of the arms, the partition becomes a positive-definite correlation matrix $\Sigma$ through an RBF kernel on cluster ranks, and the per-round posterior sample is drawn with covariance $\Sigma$ while the Beta posteriors are updated from real rewards only, so the LLM shapes how the sampler moves, not what it believes. We give a self-contained Bayesian regret bound for the idealized Gaussian sampler whose information gain splits into a $K\log T$ term from the $K$-cluster structure and a ridge term that grows to $d\log T$: the $\sqrt{d/K}$ improvement over independent sampling is a finite-horizon transient, exact only as the within-cluster correlation tends to one. The correlated sampler reduces regret by 19% over CTS on 16 synthetic Bernoulli families at $T=2{,}500$ (6-7% at $T=25{,}000$ with data-adaptive kernels) and by 41% on the Microsoft MIND-small news benchmark ($d=200$ real articles), while pseudo-observation warm starts give nothing. An LLM-free ablation with a simulated oracle of controlled quality shows that on unstructured instances the gain is a property of the kernel shape (a random partition, or a plain tempering of the sampling noise, reproduces it), while belief injection at matched oracle quality never helps.

Explicitly eliciting reasoning substantially improves LLM performance. Existing approaches require a predefined characterization of reasoning, whether through CoT prompt design, contrastive CoT directions, or via SAE derived reasoning features. For mathematical reasoning with verifiable answers, we show that a much simpler signal suffices, which is the correctness of the model's own direct answer attempts. This signal yields a latent direction that elicits reasoning. This direction is decodable within the activations of most attention heads, but only a small subset of them can be effectively intervened. We introduce COMPASS, an inference-time steering method that identifies these heads using a logit-space attribution score and steers their activations along the correctness direction, requiring only per-head activation statistics. Across three model families and multiple math benchmarks, COMPASS outperforms the activation-steering baselines we compare against, improves GSM8K accuracy by 16 percentage points on average, and approaches CoT accuracy with 20-70% fewer generated tokens. Interventions transfer without re-fitting to unseen benchmarks, and ablations show that both the correctness direction and the small set of heads carrying it are necessary, with the effect concentrated in remarkably few heads.

We give a deterministic reduction for maintaining simultaneous approximate single-source distance estimates in online decremental directed graphs. For positive integer weights in $[1,W]$, the algorithm maintains an explicit array of integer $(1+\varepsilon)$ upper estimates, identifies unreachable vertices exactly, and answers numerical queries in constant time. Let $N=m+n$, $S=N+\lceil1/\varepsilon\rceil$, and assume $\log W=\operatorname{polylog}(S)$. Initialization and all updates take $O(\Delta)+(N+\Delta_{\mathrm{eff}}+\mathcal{L}/\varepsilon)S^{o(1)}$ time, where $\Delta$ counts raw updates, $\Delta_{\mathrm{eff}}=O(m(1+\log W)/\varepsilon)$ counts filtered updates, and $\mathcal{L}$ measures finite distance growth weighted by the current indegrees of the reachable subgraph. Removed edges and vertices that become unreachable incur no subsequent growth charge. Consequently, the worst-case bound is $O(\Delta)+(N/\varepsilon)S^{o(1)}$, which is almost linear for subpolynomial inverse accuracy. The reduction uses the dynamic minimum-ratio cut and exact reachability algorithms of van den Brand et al. (FOCS 2024). A degree-weighted clipped logarithmic potential turns constant relative cut balance into accuracy at every target; a successor cut then restores strict feasibility. A distance warm start and the energy released by decreasing return coefficients yield the refined growth bound. Return mass also amplifies the accuracy of the existing change detector. The public estimates can be monotone, with $O(n+\mathcal{J}/\varepsilon)$ array writes for an unweighted finite-growth measure $\mathcal{J}\le\mathcal{L}$. Execution uses rational arithmetic. A separate reduction extends the worst-case guarantee to nonnegative integer weights. The algorithm maintains numerical estimates and does not provide fast path reporting.

Multimodal systems often encode every available input, even when a subset suffices for prediction. Adaptive acquisition can reduce this cost by using predictions from incrementally fused evidence to decide which modality to encode next and when to stop. However, sequential fusion makes these predictions order-dependent, so decisions based on them may need to distinguish factorially many histories of the same acquired set. We introduce SemARC, which couples a Sequential Modality Aggregator (SeMA) with an Adaptive Runtime Controller (ARC) and uses acquired evidence to select each modality before its encoder runs. SeMA executes only selected encoder and fusion branches, updates a fixed-size state, and predicts after each acquisition without recomputing earlier branches. We supervise every acquisition prefix under randomized modality subsets and orders to encourage consistent predictions across acquisition orders. ARC combines a set-dependent marginal-utility prior with residual fitted-Q learning to select the next available modality or stop, without inspecting unacquired inputs or retaining acquisition order. Across six multimodal classification datasets and eleven baselines, SemARC achieves 3.2% higher macro-F1 and 61.4% lower total inference GFLOPs on average relative to each dataset's most accurate baseline. End-to-end latency falls by 44.0% across GPU and CPU and by 47.2% on Android INT8 relative to the fastest measured baseline, on average. Under varying runtime modality missingness, SemARC still skips available modalities, matching or exceeding the best baseline macro-F1 in 21 of 24 conditions with 14.8% lower total GFLOPs on average. SemARC thus offers a practical path toward efficient multimodal inference across heterogeneous devices.

We give an exact algorithm for continuous separable convex quadratic programming with an integer equality matrix and nonnegative variables. The number of rational arithmetic operations and comparisons is polynomial in the dimensions and the encoding length of the constraint matrix, independently of the right-hand side, linear costs, and all nonnegative quadratic weights. Intermediate rational encodings are polynomial in the complete input. This yields a strongly polynomial algorithm for every integer matrix class with polynomially bounded entry encodings, including arbitrary matrices with entries in $\{0,\pm1\}$. The local solver compresses a quadratic objective on an integer box and solves only the compressed continuous problem. Two-sided proximity on boxes with rational bounds transfers its solution to a nearby original optimum; integer optima are used only in the proof. A single linear program supplies an error bound and a dual potential, while tight-coordinate revelation and fixed feasible anchors ensure termination and control encoding lengths. An explicit lifting extends the result to convex piecewise quadratic functions of signed integer affine features, with operation counts independent of breakpoints and objective coefficients. Applications include continuous multicommodity flow with shared capacities and piecewise quadratic costs of aggregate congestion.

While instruction-guided editors built on diffusion transformers can turn a personal photo into a convincing forgery with a single sentence, protective perturbations aim to make such edits fail before the photo is ever shared. The threat is most acute for the photos people actually share, which have 10 megapixels or more and which current editors first downscale by a factor that the owner does not know. Existing protections are computed at one resolution or averaged over random scales, and they lose most of their effect once the photo is downscaled. They implicitly assume that the editor sees the scale for which they were optimized, or that protecting the average scale protects every scale. We propose SRIM, a protection that remains effective at every scale an editor may use, without knowing its preprocessing and at the same computational cost as standard expectation over transformation. SRIM covers the range of scales implied by the photo size with anchor scales and spends each optimization step on the scale where the protection is currently weakest. A spectral analysis shows that this worst-case protection degrades only logarithmically with the width of the range, a value that averaging over scales cannot reach. Extensive experiments on full-resolution photos and popular instruction-guided editors validate that SRIM gives the strongest protection, outperforming existing methods even with a quarter of their $\ell_\infty$ budget, remaining the strongest under input purification.

Inserting objects into existing 3D scenes requires more than selecting a plausible location: the inserted object must also fit local geometry while preserving semantic intent and physical plausibility. Although recent Vision-Language Models (VLMs) and generative models enable semantic reasoning and visual content creation, they offer limited 3D grounding and geometric control when an inserted object must fit into constrained local spaces. We introduce ElasticFit, a VLM-guided framework for fit-aware object insertion centered on a novel scene-grounded representation. Given a language instruction and rendered scene observations, ElasticFit infers structured fitting cues that specify where the object should be grounded, what volume it should occupy, how it should be oriented, and its adaptation mode (rigid placement, uniform scaling, or elastic fitting). These cues convert high-level VLM reasoning into explicit 3D constraints that condition object generation and guide downstream geometric fitting. ElasticFit then generates a scene-conditioned object prior, reconstructs it in 3D, and refines the mesh through mode-specific fitting while enforcing collision avoidance, contact consistency, and physical grounding. In fixed-asset baseline comparisons, ElasticFit improves spatial relation success from 50.8% to 69.7% and support success from 48.3% to 91.7% over the strongest baseline, while providing novel support for generative "make-it-fit" insertions in complex scenarios.

Organizations make claims about their AI agents: a person approves every external email, every action is logged, an evaluation shows the agent is safe to deploy. Article 12 of the EU AI Act requires high-risk systems to allow the automatic recording of events but does not say which records settle a given claim. The position is one sentence: to be checked, a claim about an agent must first name its policy, its scope, the records that would settle it, and who writes them. Adapting the preconditions of an assurance engagement, we call a claim auditable when these elements and a decision rule are fixed before any verdict and the records are obtainable. This extends the Policy Checkability dimension of our Auditable Agents framework from single actions to claims. Agents add three conditions: coverage by an independent record, authorization bound to each action's arguments, and completeness beyond integrity. Under an explicit model, we prove that support is impossible without each wherever its hypotheses hold. A claim-check table applies the method to six common claims, anchored in current NIST, IETF, and OWASP drafts. A worked case follows one claim through five evidence states. We close with a practice box and steps for operators, buyers, auditors, and standard setters.

Hypergraphs offer a natural framework for modeling networked data, where dependencies among entities are governed by higher-order interactions. While hypergraph learning methods such as hypergraph neural networks have demonstrated remarkable predictive performance, most existing approaches rely on black-box message-passing architectures, making it difficult to disentangle the contributions of node attributes and higher-order structural information. To address this challenge, we introduce the hypergraph neural additive network (HGNAN), an inherently interpretable framework for learning on hypergraph-structured data. HGNAN extends classical neural additive models to higher-order relational data by integrating feature-wise nonlinear decomposition with hypergraph-aware structural aggregation, enabling transparent prediction for both node- and hyperedge-level tasks. Extensive experiments on benchmark datasets demonstrate that HGNAN achieves performance comparable with state-of-the-art hypergraph learning methods while providing intrinsic and meaningful interpretability.

Fine-grained mixed-precision quantization promises efficient large language model inference, but local precision choices can conflict with regular GPU storage and computation units. This precision-boundary mismatch limits the translation of compression into practical acceleration. We introduce AlignQuant, a post-training quantization method that uses GPU-compatible two-dimensional weight tiles as the common unit of precision allocation, compact storage, and execution. This shared partition lets precision follow sensitivity within output channels. Joint prefill/decode calibration scores precision reductions using projection-output perturbations weighted by language-model loss gradients under quantized activations. Phase-normalized scores prioritize higher precision for tiles important to either phase under a model-wide weight-storage budget. Each tile stores one selected representation, while phase-specialized kernels reuse the packed model and expand lower-bit weights for INT8 computation with 8-bit activations. Across four LLMs spanning 3B to 14B parameters, AlignQuant achieves up to $2.50\times$ generation speedup over BF16 while preserving model quality. Evaluations further cover three GPUs and contexts up to 64K tokens. These results show that local precision flexibility and regular GPU execution can coexist through a shared tile unit. The implementation is available at https://github.com/HanzhiZhang-Ulrica/AlignQuant.

A common assumption to simplify the problem of controlling a multi-UAV slung load system (MUSLS) is that the flexible cables can be modeled as massless rigid rods. In this work, we propose an alternative Euler-Newton derived dynamical model which uses a series of rigid links to model the flexible cables. The model is specifically designed to allow efficient simulation using Featherstone's articulated body algorithm. We perform real-world validation of this model on gentle, aggressive, and tension-engagement maneuvers and run a parameter sweep to determine the number of links, joint damping, and joint friction to achieve the greatest model fidelity. The model closely matches real-world flight data with mean load translation errors below 132 mm (5.5% of the cable length) and orientation errors below 11.4 degrees. We make the real-world flight data publicly available for the development of future cable models.

We study the one-pass streaming complexity of CSPs over a fixed finite relation $R$ under three natural objectives: $\textsf{SAT}(R)$, deciding whether all constraints can be satisfied; $\textsf{Min}$-$\textsf{CSP}(R)$, approximately minimizing the number of unsatisfied constraints; and $\textsf{Exact-CSP}(R)$, exactly computing the maximum number of satisfied constraints. Sharma and Velusamy (ESA 2026) characterized the streaming complexity of satisfiability for CSPs with literals using the non-redundancy parameter $\textsf{NRD}_n(R)$, which roughly measures the largest instance in which every constraint is independently necessary. For the most general setting without literals, they obtained the corresponding characterization for Boolean relations and showed obstacles to extending their techniques to larger domains. Kol, Paramonov, Saxena, and Yu (ITCS 2023) similarly characterized the streaming complexity of $\textsf{Exact-CSP}(R)$ for Boolean CSPs with literals in terms of the degree deg$(R)$ of the relation when written as a polynomial, while without literals, the corresponding result was known only for $\textsf{Max-Cut}$. For all three of the objectives we study, we give tight characterizations for unweighted CSP instances over arbitrary finite domains. The streaming algorithms we provide are deterministic, while the corresponding bounds are tight up to polylogarithmic factors, even against randomized algorithms. The central idea in all three results is a reduction of $R$ to its core, a canonical subrelation of $R$ which admits gadgets that allow for the hard-pinning (or fixing) of variables.

Accurate forecasting of pathological outcomes is a central problem in psychology. To do so, psychologists often collect intensive longitudinal data. However, in such studies, the desire to acquire a large number of variables for the sake of accurate prediction is often counteracted by the need to minimize participant burden. Acquiring more variables per occasion can yield better predictions, but having too many acquisitions increase the risk of non-response and attrition. Longitudinal Active Feature Acquisition (LAFA) is a principled approach to resolve this conundrum. Instead of requiring responses to every item at every acquisition occasion, LAFA produces a policy that seeks to optimally select dynamic subsets of items to be acquired at each timepoint while preserving our ability to forecast a specific outcome. However, existing LAFA methods are mostly based on Neural Networks (NN) that are difficult to interpret in practice. In this work, we introduce a tree distillation method for learning an interpretable policy from NN-based LAFA networks. We validated our method through both a simulation and an empirical EMA dataset on forecasting daily alcohol consumption. In both cases, we find that we can meaningfully reduce the number of items acquired at each occasion with minimal loss in accuracy. Networks (NN) that are difficult to interpret in practice. In this work, we introduce a tree distillation method for learning an interpretable policy from NN-based LAFA networks. We validated our method through both a simulation and an empirical EMA dataset on forecasting daily alcohol consumption. In both cases, we find that we can meaningfully reduce the number of items acquired at each occasion with minimal loss in accuracy.

Autoresearch agents tackle open-ended problems by repeatedly proposing candidate solutions, evaluating them, and using feedback to guide subsequent experiments. We show that independent runs of the same agent on the same task often plateau at substantially different scores, with gaps that persist even after considerable additional compute. Embedding their candidate artifacts by functional similarity provides further evidence that trajectories remain in localized regions of the solution space, which we call idea basins. To help agents escape these basins, we study a simple periodic intervention, fork-and-flush. Our method forks the agent into parallel trajectories, each inheriting the accumulated workspace but starting with a fresh chat context. After running each trajectory for a fixed horizon, the agent continues from the highest-scoring one. Across 13 long-horizon research and engineering tasks, with individual agent runs lasting up to several days, fork-and-flush outperformed the single-run and best-of-N baselines by a relative improvement of 66.0% and 44.4%, respectively, on the min-max normalized average score under an equal compute budget.

Large language models (LLMs) have become increasingly capable of generating code. However, achieving stronger code-generation performance still often relies on costly model adaptation, i.e., fine-tuning pretrained model parameters. Prior studies have shown correspondence between human code processing and neural models' attention or internal computation. Human-aligned learning approaches use cognitive signals to guide training, but typically adapt a large portion of the model, leaving training costs largely unchanged. Human cognitive signals may indicate not only what the model must learn from, but also where adaptation is most useful. We investigate whether human responses during code reading correspond to code-model behavior and can guide selective adaptation without sacrificing performance. We present CogAdapt, a cognition-informed framework for task-dependent sparse adaptation of code models. CogAdapt first learns transferable program-level and token-level priors from human Electroencephalography (EEG) and attention data, then combines these priors with the frozen model's response to each coding task to determine how much adaptation to allocate and which transformer blocks should receive updates. During fine-tuning, only the selected blocks are updated, while no new human recordings are required for inference. Across Qwen and GLM, we find consistent correspondence between human reading behavior and Mixture-of-Experts (MoE) computation. CogAdapt achieves the best pass@1 across both LiveCodeBench and BigCodeBench, including gains of 10.86 and 6.29 percentage points over matched regular fine-tuning on LiveCodeBench, while reducing gradient-eligible adaptation parameters by 86.21-87.21%. These results suggest that human comprehension signals can provide useful guidance for making code-model adaptation both more selective and more effective.

A GUI agent decides which action to take and where to take it; we ask how a small grounding model should receive the action type. Fine-tuning Qwen2-VL-2B with LoRA on Android in the Wild, we compare a flat baseline with five ways of supplying the type under matched data, compute, and decoding: an auxiliary loss, a hard-routed action word, an additive learned embedding, a prepended learned token, and the type written into the prompt. With five seeds, an episode-clustered bootstrap, and seed-level paired tests, the ranking on a mixed stream is clear: the auxiliary loss, the additive embedding, and the prompt word each gain five to seven hit@0.10 points over the baseline, while hard routing and the prepended token are not distinguishable from it. Much of that gain is protection from a preprocessing choice of ours rather than a spatial prior. Our serializer clamps the off-screen touch point AITW records for type events to the origin; that class degrades the baseline's click grounding, and removing it lifts the baseline by nearly seven points, after which no mechanism's hit rate beats it and the intervals exclude a two-point effect, though the auxiliary loss still shortens the average miss; on a stream of taps and swipes none helps. Whether this generalizes beyond one serialization is open. For deployment, the pipeline's margin over the baseline with predicted rather than gold types is not established (+0.016, 95% interval [-0.017, +0.052]), and a wrong type collapses every model conditioned at inference. The prepended token does not help at the shared learning rate, where its rows barely move from initialization; trained ten times faster it reaches the level of the other three, with a margin three seeds do not establish. We also document a silent failure: injecting conditioning through inputs_embeds makes Qwen2-VL fall back to 1-D positions for image tokens, costing nine points.

Large language models (LLMs) have become the backbone of modern AI applications, but pose significant challenges for efficient inference. Their autoregressive generation divides execution into two phases: prefill, dominated by large GEMMs, and decoding, dominated by small GEMVs. Modern serving systems further introduce complexity through continuous batching and prefill-decoding disaggregation, leading to dynamic workloads and phase separation. However, existing accelerators remain poorly aligned with these system-level behaviors, resulting in inefficiencies in LLM serving. In this work, we present DynaCore, a unified architecture for efficient LLM serving via system-architecture co-design. We observe that the compute tile a systolic array executes, its Minimum Efficient Unit (MEU), spans all three GEMM dimensions. DynaCore reshapes the MEU along all three: spatially it trades array width against height asymmetrically, raising weight delivery while leaving the input path untouched, and temporally Split-K maps the reduction onto the array, folding partial sums through the interconnect the array already has. To exploit phase separation, we further propose disaggregated quantization, applying dual-side quantization to prefill and weight-only quantization to decoding, with an inner-product mixed-precision datapath that keeps output width invariant to precision. A runtime scheduling framework then selects an MEU per batch. Evaluation with real-world serving traces shows that DynaCore substantially reduces service-level latency over quantization and reconfigurable accelerators, improving TTFT by 3.50x and 2.97x and TPOT by 36.55x and 8.02x, respectively.

Artifact removal routinely precedes the classification of electrodermal activity (EDA), on the assumption that a cleaner signal supports a better decision. We tested this assumption in a virtual-reality (VR) balance-disturbance task. A residual gating network was trained on a benchmark with expert-corrected EDA, frozen, and applied to VR recordings, where raw and gated signals were classified by five published time-series methods under identical leave-one-participant-out evaluation. On the benchmark the gate detected artifacts well (median record AUROC 0.94) and reduced error inside artifact regions by 17.8%. In the VR task it did not improve classification. Changes in balanced accuracy ranged from -1.35 to +0.93 percentage points, no classifier improved and two lost accuracy, and all five were equivalent to raw input within +/- 3.32 points. The benefit was lost between waveform and decision. The correction that lowered waveform error also reduced skin conductance response detection in all 43 benchmark records. Processing left 92.8% of predictions unchanged, and the predictions it did change were corrected and corrupted at similar rates. The VR recordings also carried little contamination (an estimated 4.6% of samples), and even perfect localization of deliberately injected artifacts recovered only 3.3 points in the most sensitive classifier. A pooled association between artifact level and accuracy (11.3 points) disappeared within participants (0.1 points), showing how differences between people can make cleaning look useful. Preprocessing should be judged by the decision it supports, against an unprocessed arm.

For retrieving prior cases in Indian legal judgments, the problem involves distinguishing relevant legal facts from mere lexical similarities because a prior case that shares a statute with the query judgment is not necessarily relevant. In this paper, we describe an empirical evaluation of three consecutive designs of retrieval systems for the IL PCR(Indian Legal Prior Case Retrieval) task. We demonstrate that the combination of the rhetorical roles (Fact, Ratio Of The Decision, Precedent, Argument, Statute) is better than either using individual roles or performing full text retrieval, that statute similarity is not discriminative, and that a legal entailment reranker with training data produced by an LLM is much worse in terms of official evaluation than its internal validation score. Two rankers, with excellent internal MRR up to 0.97, performed poorly in terms of official evaluation (MRR as low as 0.14). Thus, we propose a new design with query disjoint splitting and frozen validation fusion. The result is a four stage pipeline with Micro F1 0.2549, MRR 0.6081, and nDCG@10 0.4187.

Reconfigurable Intelligent Surfaces (RIS), also known as programmable metasurfaces, have emerged as a promising technology for next-generation wireless communication systems by enabling programmable control over the propagation environment. Through dynamic manipulation of the phase, amplitude, and polarization of electromagnetic waves, RIS introduces new degrees of freedom for enhancing coverage, spectral efficiency, energy efficiency, and interference management. This article reviews recent advances in RIS-enabled wireless communications, focusing on RIS architectures, emerging design and analysis techniques, and their impact on system-level performance. Key challenges related to scalability, hardware non-idealities, channel modeling, and control overhead are discussed, along with promising solutions and future research directions toward practical RIS deployment in beyond-5G and 6G networks.

When does a second artificial intelligence (AI) component improve a primary network-fraud decision rather than add operational burden? We study this question through a role-aware Decider-Supervisor (DS) framework with blockchain auditability, evaluating four directional configurations that combine centralised machine learning, a Federated Averaging (FedAvg)-trained federated meta-model, and Base or Quantized Low-Rank Adaptation (QLoRA) large language model variants. The analysis compares primary-only and supervised decisions using non-hard fraud performance, intervention burden, conditional calibration, traffic-mix and Review-capacity sensitivity, dependability tests, and blockchain lifecycle controls. The deterministic hard gate resolves 89.994% of fraudulent requests, leaving the non-hard population as the main AI decision setting. Conditional validation calibration does not produce a consistently transferable supervisory advantage on deployment replay. DS-3 QLoRA is the least disruptive supervised configuration, but it still underperforms its primary FedAvg stage in F1 and total errors. Across 36 reweighted traffic mixtures, supervision reduces total errors only for DS-4 Base in two extreme high-fraud scenarios. Blockchain tests support digest verification, tamper detection, authorisation, single-use review resolution, and post-finalisation integrity, while exposing a pre-finalisation single-write limitation. The results show that the value of AI supervision depends on role assignment, calibration, escalation policy, traffic composition, and lifecycle controls rather than on the presence of a second model alone.

We study decentralized multi-agent target search where homogeneous agents communicate intermittently at Poisson-distributed times. Standard unconditional belief fusion wastes communication opportunities by synchronizing agents during high-entropy exploration, when diverse independent beliefs provide better coverage than a premature consensus. We introduce entropy-gated belief coordination, in which agents skip fusion while their collective entropy ratio exceeds a threshold~\(\theta\) and merge only during exploitation, consistent with the bifurcation structure of nonlinear opinion dynamics and the submodular structure of the per-step information gain objective. We further derive \(\Istep(\alpha,\beta)\), the expected mutual information per observation step between two agents' binary sensors, as an interpretable, communication-free measure of sensor informativeness that motivates the gating design and guides system-level analysis. Experiments across 103{,}680 trials (nine grid sizes up to \(100{\times}100\), Poisson communication timing, four target movement patterns) show that the Entropy-Gated Trust-Decay Planner (EG-TDP), which adds a detection-probability planner switch in exploitation mode, achieves mean belief quality \(\bar{Q}=0.300\) (mean belief mass at the true target cell, averaged across all trials and steps), a \(58.9\%\) gain over arithmetic mean and a \(37.4\%\) gain over visit-weighted fusion. On representative configurations, EG-TDP also outperforms a joint-Bayesian reference that uses all agents'~observations at every step, despite operating under random intermittent contact only.

Reducing nonstationarity in a persistent time series entails deciding how much of its temporal dependence to remove. In finance, fractional differencing is often tuned using the Augmented Dickey–Fuller (ADF) test, limiting the search to a one-parameter family of lag profiles and addressing input preservation only indirectly. We propose StaFIR, a causal finite-impulse-response filter with a learned nonnegative mixture of exponential lag profiles. Its convex learning objective balances empirical stationarity with similarity to the input. We evaluate StaFIR on ARFIMA–GARCH controlled settings and rolling financial series, including a realized-volatility forecasting task. The experiments show that StaFIR adjusts its filtering strength to persistence while limiting unnecessary transformation in stationary regimes. In downstream forecasting, there is no clear accuracy difference from fixed half-order differencing, while StaFIR achieves higher measured similarity to the raw signal. A complementary direct forecasting experiment finds that greater input similarity is associated with smaller forecasting penalties, although the raw representation remains stronger.

Adaptive gait biofeedback may support repeated practice in chronic ankle instability, but its evaluation must address model performance and human response. We evaluated a temporal convolutional classifier on protocol-defined, angle-derived GOOD/BAD gait-cycle labels using participant-held-out leave-one-subject-out (LOSO) cross-validation in 20 participants. Seven participants in the adaptive-intervention group completed nine sessions over three weeks, with one motion-capture recording analyzed per session. Models updated after failed sessions were compared offline with their parent models on the same-session validation subset used for candidate selection and the first subsequent adaptive-session recording. Frontal-plane ankle angle was compared between the adaptive group and 10 sequentially enrolled controls at Baseline, Post, and 7-day Retention. Across 20 held-out folds, mean fold-level area under the receiver operating characteristic curve (AUROC) was 0.948, sensitivity for angle-threshold-exceeding BAD cycles was 0.941, and specificity for angle-threshold-meeting GOOD cycles was 0.366. Mean BAD-class F1 was higher in candidate models by 0.187 on the same-session subset and 0.118 on the first subsequent recording. At Post, the adaptive group had a baseline-adjusted frontal-plane ankle angle 5.168 degrees lower than controls (95% confidence interval, 1.766-8.569 degrees lower); the Retention contrast was uncertain. These findings characterize population-model discrimination and offline participant-specific updating during repeated biofeedback use, alongside a nonrandomized Post frontal-plane ankle angle association. They do not establish independent clinical gait classification or a causal benefit of updating.

Riot Games, which runs the League of Legends World Championship (Worlds), also changes the game shortly before each tournament. This concentration of authority raises a governance question: Who scrutinizes how the benefits and burdens of late rule changes are distributed? We examine all ten championships from 2016 to 2025, combining 1,097 games, 215 team-years, and LLM coding of 1,090 champion changes across 40 patches. On average, 33% of a team's prior picks were on champions with net nerfs in the patches it had not played before Worlds, and 8% were on champions with net buffs. Within a year, the median gap between the most and least affected teams was 23 percentage points for nerfed picks and 12 for buffed picks. Nerfs fell mainly on the most-contested champions and buffs on less-contested ones, as Riot's published balancing criteria anticipate. Pools concentrated on the prevailing meta therefore had more negative net exposure, and even the winners of a year's four major regional leagues differed in net exposure about as much as four randomly drawn teams. Higher buff exposure was associated with additional opponent bans on buffed champions, although its size depends on the model. Among quarterfinalists, the most buff-exposed team of each year reached six of ten finals and won four titles. The team with the most negative net exposure reached two finals and won one. These exploratory descriptions identify unequal exposure, not causal gains or losses. The association between net exposure and winning remains uncertain, at +1.3 percentage points per standard deviation with a 95% interval from -0.9 to +3.7. We argue that scrutiny of late rulemaking should not depend on proving that a patch changed who won Worlds. Earlier patch commitments, team-level impact disclosure, independent review, and a formal opportunity to challenge decisions would make publisher authority more accountable.

Accent classifiers are typically trained with a fixed label inventory and cannot accommodate new accent categories as new data becomes available. Moreover, accented speech corpora often exhibit substantial class imbalance and/or domain shift due to differences in recording conditions across corpora. We present AccentCL, a class-incremental learning framework for English accent classification that is robust to class imbalance and cross-corpus domain shift. AccentCL extracts multi-layer representations from a frozen Whisper-Large-v3 encoder, optimized with an imbalance-aware cross-entropy loss to reduce bias toward the majority accent classes and a domain mean alignment loss that minimizes distributional mean shift across training corpora. The label space is then expanded via replay-based continual learning, using the frozen base model for knowledge retention and an old-to-new margin loss to reduce overprediction on newly added classes. On a five-class accent classification task, AccentCL achieves 77.1% balanced accuracy and a 76.9% macro-averaged F1 score. We further evaluate the model's ability to incrementally incorporate two new accent categories: Spanish-accented and Chinese-accented English. When adding Spanish-accented English to the pretrained model, AccentCL attains an F1 of 83.3% on the new class while retaining 77.3% balanced accuracy on the base classes. When subsequently adding Chinese-accented English, it achieves 61.8% F1 on the new class while preserving 77.6% balanced accuracy on the previously learned classes. These results show that AccentCL enables robust regional accent classification while allowing new accent categories to be added without full retraining.

Field-effect transistor (FET) layouts on exfoliated two-dimensional flakes are typically drawn by hand for each flake, placing contacts and gates to match its position and outline in optical micrographs. To our knowledge, no executable benchmark tests whether language-model agents can perform this flake-specific construction reliably. We introduce 2D-FET-Bench V2, a benchmark of 128 layout tasks built from microscopy-derived flake contours, including hole-containing flakes and multi-flake tasks. Each task supplies a textual device specification and contour coordinates. An agent generates typed polygon and path operations rendered to GDSII. A deterministic verifier checks geometric and structural requirements, and a separate integrity check verifies that the supplied contours remain unchanged. Scripted reference layouts pass all 128 tasks, showing that every task is solvable. We evaluate six models and seven workflow and scaffold variants of GPT5.6-Luna, with five attempts per task. The best-performing configuration in the six-model panel, GPT5.6-Luna with ReAct-3, passes 62.3% of attempts and solves 80.5% of tasks at least once (coverage) and 43.8% in all five attempts (consistency). ReAct-3 exceeds the one-pass Plan-and-Execute by 27.0 pass@1 points at 2.46 times the tokens. An expert audit of one sampled verifier-passing layout per covered task, across five ReAct-3 configurations, accepts 56.4% to 63.5% of them. The benchmark evaluates geometric and structural FET layout construction.