Papers

This paper presents a comprehensive experimental video quality assessment in 5G-based Fixed Wireless Access (FWA) networks, leveraging a real open source 5G network testbed. Addressing the critical need for robust UHD video delivery, the study systematically investigates the impact of transmission modes (SISO and MIMO), modulation schemes (64-QAM and 256-QAM), and network load (one to five users) on both network throughput and perceived video quality. Unlike simulation-based approaches, our methodology integrates automated streaming campaigns with objective full-reference Quality of Experience (QoE) metrics (PSNR, ssim, and VMAF) computed from 4K video streams delivered via MPEG-DASH, enabling frame-level correlation between network performance and user experience. Key findings demonstrate that MIMO consistently outperforms SISO, sustaining higher throughput and superior visual quality even under increasing user demand. Specifically, MIMO with 256-QAM maintained VMAF scores above the visually lossless threshold with five concurrent users, while SISO configurations experienced significant degradation. These results provide empirical evidence for the critical role of spatial diversity and higher-order modulation in ensuring reliable UHD video delivery over FWA. The proposed reproducible framework for detailed quality assessment offers valuable insights for future optimization of 5G FWA deployments, contributing to the practical understanding and deployment strategies for high-quality video services in Open RAN environments.

Cybersickness remains one of the primary barriers to prolonged and comfortable virtual reality (VR) use, yet the temporal development of cybersickness remains poorly understood because its onset and progression vary substantially across individuals. Eye tracking provides a continuous and unobtrusive measure of user behavior, making it well suited for investigating how users respond throughout VR exposure. However, existing work has mainly focused on aggregate gaze measures or post-exposure assessments, leaving the relationship between early eye-movement dynamics and discomfort progression underexplored. In this paper, we investigate whether early eye-movement dynamics characterize the subsequent progression of cybersickness. Using trajectories of fixation duration, fixation dispersion, saccade amplitude, saccade velocity, and gaze eccentricity computed over the first one to five minutes of VR exposure, we evaluate their association with subsequent discomfort progression and compare early gaze dynamics between participants with low and high post-exposure cybersickness severity. Our results show that early fixation dispersion is consistently associated with subsequent discomfort progression, whereas no significant differences are observed between cybersickness severity groups. These findings suggest that early fixation dispersion can characterize the progression of discomfort rather than the final cybersickness severity. Overall, this work provides new insight into the temporal relationship between early eye-movement behavior and cybersickness progression, contributing to the understanding of gaze dynamics during immersive experiences.

As autonomous AI agents take on every stage of scientific inquiry, research output is expanding far beyond human review capacity. Yet scientific communication still relies on natural-language prose: an informal medium prone to ambiguity, hidden assumptions, and untracked limitations that machines cannot reliably audit. We introduce Lara, a machine-checkable language and protocol for checking and revising support for research claims. By turning research arguments into executable artifacts, Lara provides an epistemic kernel for autonomous science: it enables automated validation pipelines for research agents, lets declared bridges connect arguments across papers into an auditable network, and allows both humans and machines to recheck the standing of an encoded claim in milliseconds. In a Lara program, authors explicitly declare their claims, supporting evidence and assumptions, and known objections or limitations. A lightweight, deterministic checker adjudicates these interactions, assigning each claim a reproducible status: "justified", "defeated", "contested", or "gap", which marks a claim whose support is incomplete and locates the unanswered question. Case studies cover empirical review, a philosophical debate without measurements, and the loss of support when an assumed axiom is withdrawn. We establish the metatheory of claim checking and cross-context argument transport, and mechanize the semantic guarantees in Lean 4 (roughly 117,000 lines), leaving three arguments on paper. The audited public metatheory is "sorry"-free and uses only Lean's three standard axioms; some executable examples additionally trust native evaluation.

Translocations are genome rearrangement operations that exchange prefixes of two chromosomes. We study the non-uniform non-contiguous translocation distance problem, where every string produced during the computation remains available for reuse. Given an initial set of strings $A$ and a target set $B$, the objective is to produce all strings in $B$ using as few translocations as possible. We present the first polynomial-time approximation algorithm for this problem. For a single target string of length $n$, we obtain an $O(\log n)$-approximation, and we extend the result to arbitrary finite target sets with an $O(\log N)$-approximation, where $N$ is the total length of the targets not already present in the initial set. This resolves the approximability question for the non-uniform non-contiguous case left open by Constantin and Popa (TCS 2025).

The $k$-core of a graph is its (unique) largest subgraph with minimum degree at least $k$. For any $k \geq 3$, deciding whether a given vertex belongs to the $k$-core is a P-complete problem, meaning that it is inherently sequential and highly unlikely to admit efficient parallel algorithms, even on graphs of maximum degree $k+1$. This paper investigates alternative parameterizations of the $k$-core problem to identify conditions under which it can be placed into sub-polynomial complexity classes. We prove that the problem is in para-NC$^{2+\epsilon}$ when parameterized by treewidth, and in para-NC$^3$ when parameterized by $k$ on chordal graphs. Furthermore, we introduce a novel NC$^{3}$ algorithm for interval graphs when $k = \mathcal{O}(\lg v(G))$, which relies on an improved parameterization by pathwidth. Finally, we establish corresponding lower bounds, demonstrating that, even with these parameterizations, computing the $k$-core remains L-hard, meaning it requires at least logarithmic space. These findings explore the boundary of parallel tractability for the $k$-core problem by highlighting the graph parameters that make it inherently sequential.

Ultrafast pump-probe spectroscopy provides a powerful means to investigate electron-phonon (e-ph) interactions in strongly correlated systems. Nevertheless, the question of how different microscopic e-ph coupling mechanisms influence the lattice's nonequilibrium response has not been widely addressed. We perform time-dependent density matrix renormalization group simulations on a pumped one-dimensional Mott insulator coupled to phonons through either Holstein or Su-Schrieffer-Heeger (SSH) interactions. The electronic responses of the two models are qualitatively similar in the Mott-insulating regime and are governed by the same excitations and resonance conditions of the Hubbard model. In contrast, the lattice dynamics of the two models differ substantially. While the Hubbard-Holstein model exhibits long-lived coherent phonon oscillations when the pump resonates with the bare phonon frequency, no such oscillations are observed for the Hubbard-SSH model. These contrasting responses remain robust against changes in phonon frequency and pump parameters, suggesting that coherent lattice dynamics can provide a direct nonequilibrium probe of the microscopic form of e-ph coupling.

Biological torque control directly maps an estimated human joint moment to exoskeleton assistance, providing a task-agnostic strategy for supporting diverse locomotor activities. However, it remains unclear whether a fixed state-to-torque mapping provides effective assistance across biomechanically distinct tasks. We examined how assistance delay affected hip exoskeleton performance during level-ground (LG), ramp-ascent (RA), and ramp-descent (RD) walking. Eight participants completed a zero-torque baseline condition and five active assistance conditions with delays ranging from 40 to 320 ms. Across tasks and active delays, assistance reduced net metabolic rate by 5.24%, positive biological hip joint work by 5.86%, and total lower-limb positive joint work by 1.68% (all p < 0.05). Assistance delay affected both joint-work outcomes (both p < 0.001) but not net metabolic rate. Mechanical unloading generally decreased with increasing delay, whereas metabolic benefits remained comparatively stable. Relative to the zero-torque condition, net metabolic rate decreased by 9.75% during LG and 7.20% during RA but increased by 1.23% during RD. We did not detect task-dependent differences in the delay response. Our findings indicate that biological torque mappings should be evaluated based on the target outcome and mechanical role of the assisted joint, and that predominantly positive-power assistance may not generalize to negative-work-dominant locomotion without modification.

Nirenberg's theorem gives interior $C^{2,\alpha}$ regularity for continuous viscosity solutions of fully nonlinear uniformly elliptic equations $F(D^2u)=0$ in dimension two. In dimensions five and higher, the work of Nadirashvili–Tkachev–Vlăduţ shows that such solutions need not be $C^2$. We construct a family of $C^{1,1}\setminus C^2$ functions $\{w_m\}_{m\geq3}$ and uniformly elliptic operators $F_m$ such that $F_m(D^2 w_m)=0$ in $\mathbb{R}^{d_m}$, where $d_m=m(m+1)/2-1$. These solutions are given by $\operatorname{tr}(X^3)/\sqrt{\operatorname{tr}(X^2)}$ on the space of traceless real symmetric $m\times m$ matrices. When $m=3,4$, this family recovers the known singular solutions associated with the five-dimensional Cartan isoparametric cubic and the nine-dimensional Hsiang minimal cubic, respectively. Our construction treats all $m\geq3$ in a unified way by direct analysis on traceless symmetric matrices. We also prove a geometric criterion for pointwise $C^{2,\beta}$ regularity of continuous viscosity solutions $u$ of uniformly elliptic equations $F(D^2u)=0$ in every dimension $n\geq 3$. Suppose that $n-2$ smooth hypersurfaces pass through a point $p$, each carrying a smooth vector field along which the directional derivative of $u$ is constant on that hypersurface. If these vector fields are linearly independent at $p$, then $u$ is twice differentiable at $p$ and admits a pointwise $C^{2,\beta}$ expansion for some $\beta\in(0,1)$. Our criterion is quantitative and yields interior $C^{2,\beta}$ regularity under uniform geometric hypotheses. As an application, we extend Nirenberg's theorem to a general class of solutions in dimensions three and higher. As a consequence, we establish interior $C^{2,\beta}$ regularity for solutions of Dirichlet problems with certain group symmetries. This extends a result of Nadirashvili–Vlăduţ for axially symmetric problems.

Vision-Language Models (VLMs) enable edge devices like unmanned aerial vehicles (UAVs) to interpret visual observations and reason about complex environments using natural-language instructions. However, their practical deployment remains challenging as onboard inference is constrained by limited computational, memory, and energy resources, whereas cloud-based inference introduces communication latency, bandwidth overhead, and dependence on network connectivity. To address these limitations, split computing offers a promising alternative by partitioning VLM inference between the resource-constrained UAVs and more capable remote servers. However, the performance trade-offs among fully onboard, cloud-based, and split-computing architectures for lightweight VLMs have not yet been systematically profiled. This paper benchmarks these three deployment paradigms using SmolVLM-256M as a representative lightweight VLM. We quantify their inference latency, computational resource utilization, communication overhead, and energy consumption across varying image resolutions and network conditions. Our results show that no deployment strategy is universally optimal; instead, the preferred strategy depends on the interaction between network conditions and input image resolution.

Dense virtual environments present significant challenges for object selection and manipulation, motivating the development of novel interaction techniques. This paper presents RodCast as an exploratory case study to investigate explicit trajectory visualization for interaction in dense virtual environments. We conducted a within-subjects user study comparing RodCast with Go-Go Hand and FlowerCone across three representative interaction tasks using objective performance measures, subjective evaluations, and qualitative feedback. The results revealed that the proposed implementation incurred performance costs on more demanding manipulation tasks, while qualitative feedback suggested that participants attributed these challenges to the complexity of the bimanual control scheme. Additionally, subjective evaluations and qualitative feedback identified benefits in spatial awareness and target accessibility that were not fully reflected by conventional performance measures. Together, these findings highlight the importance of control simplicity in interaction technique design and emphasize that incorporating qualitative feedback into the evaluation process is essential for distinguishing implementation limitations from the potential of the interaction concept. Collectively, these lessons provide guidance for the design and evaluation of future interaction techniques.

Third-party evaluations for frontier AI have mostly tested models through external interfaces before deployment. But the risks from frontier AI models depend on how their developers use and govern them internally. Recently, CEOs of frontier AI companies have committed to hosting embedded assessments. These assessments would give independent evaluators employee-like access to a developer's internal systems, staff, and documentation. First, we argue that this can enable deeper and more flexible assessments of risks that depend on internal systems and practices, while providing access under stronger security controls. Then, we examine seven design questions about scope, information gathering, duration, timing, terms of engagement, disclosure, and escalation. We recommend that frontier AI developers begin hosting embedded assessments now, covering at least three areas central to managing risks from internal AI use: internal agent monitoring, internal agent security controls and permissions, and model alignment. To enable meaningful third-party scrutiny, assessments should be continuous, evaluators should publish detailed reports at least quarterly, and clear escalation mechanisms should be established. These recommendations are intended as a starting point, with further steps needed to realize the full potential of embedded assessments.

The J-region asymptotic giant branch (JAGB) method for distance measurements uses the near uniform luminosity of carbon-rich AGB stars as a standard candle. Its application rests on an assumed but rarely examined premise: the stars contained in the JAGB color-magnitude selection box are primarily carbon-rich AGB stars. We investigate the purity of the photometric selection in the Milky Way with an optical spectroscopic survey (~3590-7150 $\mathring{A}$) called CHASE (Carbon stars, Hubble tension, And Spectroscopic Exploration of JAGB populations) using the Shane 3 m telescope at Lick Observatory. These stars are classified using molecular bandhead step indices, which recover 100% of known classifications. We report the observations of 76 stars (54 within the standard JAGB selection box of 1.5 < J - K < 2.0 mag) and find only 13 of 54 stars are carbon-rich, a 76$\pm$6 oxygen-rich contamination under blind photometric selection. This contamination remains high (71%) even within $\pm$0.4 mag of the JAGB peak itself, showing it is not confined to the edges of the selection box. This inhomogeneity may explain why recent literature find field-to-field and methodological variations of 0.1-0.2 mag that now dominate its systematic error budget. These variations could be caused, in part, by inhomogeneous stellar populations that consist of differing spectral types, and hence differing mean magnitudes.

Classifier-free Guidance (CFG) is widely adopted in text-to-speech (TTS) systems to enhance generation quality and conditioning fidelity by interpolating between conditioned and unconditioned predictions. A common unconditional technique is to use an empty representation, in the form of a fixed null vector. In this work, we propose replacing this representation with a learnable unconditional embedding, optimized to represent a meaningful unconditional state. Objective and subjective evaluations demonstrate that learnable null embeddings consistently outperform fixed null embeddings across speaker similarity, speech stability, and expressiveness, while exhibiting greater robustness to larger guidance scales. We further show that learning a distinct unconditional embedding for each of the TTS conditioning modalities allows fine-grained control over speaker and text guidance, showcasing the trade-off between similarity and quality, and stability and expressiveness in the generated speech.

We study strategic disclosure of the action space in principal-agent contracting, where an agent selects a disclosed action set to shape the principal's perception of her capabilities before contract design. Unaware of the strategic disclosure, the principal designs a revenue-optimal contract as if the disclosed action set were complete and accurate. We consider two variants distinguished by cost verifiability. When costs are unverifiable, the agent can extract the entire first-best surplus, leaving the principal with zero revenue. When costs are verifiable, we characterize the agent's optimal disclosure strategy in binary-outcome settings and, more generally, when the principal is restricted to linear contracts, reducing the agent's problem to a two-variable convex optimization problem. We prove that the agent can secure utility of at least a $1/e$ fraction of the first-best surplus, which also yields a $1/e$ welfare guarantee under optimal disclosure. While the principal's revenue can be arbitrarily small compared to the first-best surplus, when the ratio of maximum to minimum expected reward among non-null base actions is at most $L$, we establish a revenue guarantee of $\Theta(1/\log L)$ relative to the first-best surplus. We also compare utilities and welfare under strategic disclosure with their counterparts in the canonical model. Finally, we extend the agent's $1/e$ utility guarantee to general outcome spaces without restricting the principal to linear contracts. Our results show how strategic action-space disclosure changes the distribution of surplus while preserving a constant-factor welfare guarantee under the agent's optimal disclosure.

In this paper we present a new result that exhibits in a non-recursive fashion a primary decomposition of any monomial ideal. Our result is derived from the classical quotient-sum recursive procedure for determining a primary decomposition of any proper ideal in a commutative Noetherian ring. We also show how our result can be used as an algorithm for producing a primary decomposition directly from the exponent vectors of the minimal generators of the ideal.

Online A/B experiments are the decision standard for user engagement, but traffic and readout time limit how many conversational-AI changes can be tested. We ask whether an offline signal designed to be computable without treatment-arm user exposure agrees with the outcomes of those experiments. We contribute a reusable construction and diagnosis checklist that treats an offline proxy as a chain of three alignments: behavioral label to product outcome, learned classifier to candidate-assistant behavior, and aggregated offline signal to experiment effect. A companion evaluation protocol audits the whole composite by interval-aware decision agreement, which compares offline and online confidence intervals instead of point estimates, and by within-experiment ranking. The instantiation we evaluate comprises a fixed evaluation suite on which candidate behavior is scored, an engagement classifier trained to predict session/prompt level engagements, and a calibration layer mapping sample-level score differences to online model-level engagement deltas. We then report the audit: 489 paired offline-online contrasts (one candidate arm against its control) from 27 experiments on a deployed multi-turn assistant, spanning model checkpoints to system-prompt tuning. Our primary test uses the 113 contrasts from eight experiments that ran after the map was frozen: on these the composite reaches 81.1% F1, against 34.3% for the raw classifier score it is built on, and makes no wrong-direction calls where that raw score makes 31. Every offline prediction was computed before its experiment ran to prevent overfitting. The evidence supports using the composite to prioritize candidates before scarce experiment traffic is allocated—in our deployment of the experiment, selecting among training checkpoints and tuning system prompts.

Sheet-like structures are ubiquitous in magnetohydrodynamic systems and are known to play a crucial role in astrophysical plasma transport. Here, by applying a theoretical framework to data from numerical simulations, we demonstrate that current-sheet formation is a direct consequence of energy conversion between kinetic and magnetic fields. The strain-magnetic field organisation associated with this conversion favours the emergence of sheet-like structures, and analytical calculations show that sheet-like geometry constitutes the dominant flow configuration. We further show that the same conversion-driven reorganisation of the strain motion is responsible for the depletion of the mean Inertial energy transfer, a previously observed phenomenon whose physical origin has remained unclear. Finally, we identify experimentally accessible diagnostics through which these predictions can be tested in multi-spacecraft observations of turbulent plasmas.

The conventional spin Hall effect generates spin currents whose flow direction, spin polarization, and driving electric field are mutually perpendicular. Magnetic order lifts this symmetry restriction and enables additional time-reversal-symmetry-odd components of the spin-conductivity tensor, giving rise to the magnetic spin Hall effect (MSHE). These components also couple the generated spin polarization to the magnetic order, providing a degree of control absent in the conventional spin Hall effect. Although the MSHE has been observed in antiferromagnets, its experimental identification in conventional ferromagnets has remained elusive. Here, using a non-local lateral spin-valve geometry, we electrically identify the MSHE and its reciprocal effect in a perpendicularly magnetized Co-based multilayer. Reversal of the multilayer magnetization reverses the MSHE and magnetic inverse spin Hall signals, revealing their time-reversal-symmetry-odd character and magnetization control. By contrast, the conventional spin Hall and inverse spin Hall signals measured in the same devices remain unchanged under magnetization reversal, consistent with their time-reversal-symmetry-even character. We obtain a magnetic spin Hall angle of $\theta_{\mathrm{MSH}} = (3.8 \pm 0.6)\%$, comparable in magnitude to the spin Hall angle of heavy metals commonly used in spintronic devices, such as Pt. These results establish the MSHE as a sizable, magnetically switchable spin-charge interconversion mechanism in conventional ferromagnets.

Modern information retrieval increasingly employs both embedding and generative models to handle complex queries. However, current serving systems suffer from low throughput and poor GPU utilization because they execute these models in isolation. Coarse-grained partitioning, such as dedicating GPUs to specific tasks, fails to adapt to dynamic workloads and creates computational "bubbles". To address these, we present Orthrus, a serving system that performs heterogeneous batching within a unified inference loop. The primary challenge lies in unifying embedding and generation workloads with conflicting computational patterns while optimizing batch composition for high performance. Orthrus addresses these challenges through chunked embedding with incremental pooling and by adjusting batch composition in a workload-aware manner. Evaluation on four A100 GPUs shows that, relative to baseline deployments, Orthrus achieves 1.28$\times$–4.52$\times$ higher throughput on controlled workloads and up to 55.8% lower end-to-end p99 latency on an iterative-RAG benchmark. We release our code at https://github.com/illinoisdata/Orthrus .

We study asymptotic spectral synthesis on $\mathbb Z_N^d$ using the Fourier ratio, the quotient of the counting-measure $\ell^1$ and $\ell^2$ norms of the unitary Fourier transform. For supports of cardinality $N^{\alpha+o(1)}$, the Fourier-ratio concentration exponent satisfies $0\leq\kappa\leq\frac{\alpha}{2}$. We prove a finite-scale synthesis estimate that uses either the Fourier ratio of a function or that of its support indicator. In particular, the indicator condition applies uniformly to arbitrary weights on the support. Uniform $\ell^p$ bounds imply decay of the largest Fourier coefficient when $2\leq p<\frac{2(d-2\kappa)}{\alpha-2\kappa}$, with every finite $p\geq2$ allowed at maximal concentration. Random unions of cosets show that the finite critical exponent is sharp for every admissible pair in each prescribed ambient dimension, along prime-power moduli. We also determine the exact synthesis constant for sets with a transitive group of affine symmetries. For paraboloids this constant is an explicit divisor sum. The formula proves endpoint synthesis along powers of a fixed odd prime, although the endpoint fails along primes. Along products of distinct small primes, synthesis extends to a larger exponent. These families have the same Fourier-ratio exponent. Finally, we relate Fourier concentration to generalized Salem estimates, additive energy, Fourier algebra norms, and necessary conditions for extension estimates.

The motility of microorganisms provides a natural framework for studying active matter and its response to environmental perturbations. We experimentally investigate the trajectories of Paramecium caudatum and Artemia salina under controlled conditions. Using video microscopy, we reconstruct individual trajectories over time. Beyond baseline conditions, we analyze stressed environments where P. caudatum are exposed to a toxic agent and A. salina are placed in distilled water. To characterize the motion, we compute trajectory-based metrics, including mean square displacement, velocity distributions, persistence times, tortuosity, and radius of gyration. These observables quantify changes in motility patterns across species and conditions. To interpret the dynamics, we develop coarse-grained stochastic models parameterized directly from experimental data. P. caudatum is described within a run-and-tumble framework, including reversal events under stress. In contrast, A. salina requires a run-and-tumble-like description with persistent orientational noise, accounting for continuous directional fluctuations during runs. The models reproduce the main trajectory-level observables used for comparison, namely tortuosity and radius of gyration. This combined approach provides insight into how environmental stress modifies locomotion statistics. Our results highlight systematic differences in the statistical signatures of motion, suggesting that trajectory metrics may serve as indicators of physiological stress. This work provides an accessible framework for active matter experiments where individual-level statistics can be directly measured and modeled.

We prove that the original Barrett–Garcke–Nürnberg (BGN) method for mean curvature flow of curves converges in the manifold distance with order $h^2$ for the time step $\tau=h^2$. The key new technique is backward error analysis, which stems from the numerical analysis of ordinary differential equations. By combining the expansions of the discrete velocity in space and time, we construct an approximation flow whose interpolation satisfies the BGN method with an improved defect, of order $h^4$ with tangential part of order $h^6$. This construction also identifies the tangential velocity of the BGN method and shows that $\tau=O(h^2)$ is the correct relation between $\tau$ and $h$. With the improved defect, we then prove $H^1$ superconvergence of order $h^4$ of the numerical solution to the approximation flow using techniques of evolving surface finite element methods, which yields the convergence in the manifold distance. Numerical experiments confirm the predicted orders of defects.

In his 1957 paper on the Dirichlet problem for higher-order elliptic equations in the plane, Agmon conjectured that there should exist higher-order equations in dimensions greater than two for which a potential-theoretic approach based on multi-layer integral operators can be successfully implemented. We show that this is indeed the case for the polyharmonic operator $\Delta^m$ of arbitrary order $m\in\mathbb{N}$ in $\mathbb{R}^n$, $n\geq 2$, even in geometrically rough domains permitting singularities beyond the scope of the existing theory of higher-order elliptic boundary value problems. Central to our approach is a new genre of multi-layer singular integral operators, which play for $\Delta^m$ the role assumed by the classical harmonic double layer potential in the treatment of the Dirichlet problem for harmonic functions.

Failure-based textual knowledge distillation aims to discover gaps in a model's knowledge by examining its task errors. The distilled knowledge can be useful for the reasoning of both this model ("source model") and other models. However, this transfer of knowledge may not be stable. We define a rule atom to be a standalone rule injected into a model's textual input at inference time. A rule atom can encode transferable task knowledge or model-specific reasoning patches that can confuse other models. Also, the injected rule atoms can be misapplied to unrelated cases, causing the model to incorrectly flip its answer based on irrelevant information. Building on a pipeline that distills training examples into task-specific cheat sheets that aid model reasoning, we examine when failure-derived rules can improve these cheat sheets. Our early experiment shows rule distillation from a single model's failures underperforms the baseline cheat sheet on non-source model families. This motivates Robust Failure, Conservative Repair (RFCR), a textual distillation procedure that derives rules from failures shared across models, sharpens their application boundaries using boundary cases, and abstains when no useful rule is found. On a 400-item BIG-Bench Hard task set, RFCR improves the baseline cheat sheets from 68.50% to 71.25% (+2.75 pp; 95% CI [+1.25,+4.50]) without performance degradation on previously correct cases. Ablations and cross-model diagnostics support that accuracy gains come from both new knowledge injection and strict rule-application control.

The presence of geometric frustration in self-assemblies, stemming from a shape misfit between identical subunits, can lead to the spontaneous formation of finite-sized structures known as self-limitation. Recently proposed models of curved, colloidal shells, dubbed "curvamers", that form one-dimensional stacks have shown curvature-induced frustration to be a promising class of particle designs for engineering self-limiting assembly. However, existing models of curvamer assembly considered only cylindrically curved shells where elastic costs are derived purely from bending. Here, we study the self-limiting behavior of a more general class of curvature-frustrated colloidal shells, extending shapes to include non-Euclidean geometries such as spherical caps and saddles whose deformations can also incur stretching costs. We introduce continuum mechanical and discrete, coarse-grained models of shell stacks to study how qualitatively different intra-assembly strain gradients required for each geometry affect the buildup of super-extensive elastic costs. We find the accumulation of elastic energies in non-Euclidean shell stacks is sensitive to shell thickness via a dimensionless quantity related to the Foppl-von Karman number that characterizes the relative strength of stretching to bending. In particular, thin shells penalize stretching more causing faster elastic energy growth resulting in the suppression of self-limitation with sizes generally largest for cylindrical shells, smallest for spherical caps and intermediate for saddles. We additionally find that while thin spherical caps experience the strongest elastic penalties, the rapidly increasing rate of elastic penalties in stacks of saddles lead to a robust size-selection as inter-particle binding strength grows as well as relatively smaller dispersity in self-limiting stack size in comparison to cylindrical and spherical curvamers.

Large-scale datasets are essential for training generalist robot control policies. Collecting real-world tactile data is costly and time-consuming, motivating the use of tactile simulations. However, current tactile simulators capture only overall contact geometry and miss fine details like texture. This results in a significant domain shift between simulated and real tactile data. To address this gap, we introduce Norm2Tex, a plug-in method that augments simulations of vision-based tactile sensors with high-frequency surface details from normal map textures. By modifying the target object's depth map before a tactile simulator's rendering pipeline, Norm2Tex seamlessly integrates into different tactile simulators. We also evaluate sim-to-real transfer using material classification and a reinforcement learning task. Our results show that Norm2Tex preserves material-dependent tactile information across domains, improving texture recognition and producing material-dependent control behavior in the real world.

The one-dimensional bin packing problem (1D-BPP) is a classical NP-hard combinatorial optimization problem with applications ranging from logistics and manufacturing to cloud resource management. Although deep reinforcement learning (DRL) has become a competitive paradigm for data-driven optimization, most learned packing methods target 2D and 3D variants, and intelligent learned solvers for 1D-BPP remain scarce. In this paper, we present a novel end-to-end, size-agnostic graph reinforcement learning framework for 1D-BPP. We formulate the packing process as a Markov decision process on an item-compatibility graph, serving as a structural knowledge representation in which every action merges two partial bins that fit together. A graph neural network actor-critic policy extracts relational features from this representation and is trained through reinforcement learning and decoded by stochastic beam search, enabling a single trained model to generalize zero-shot to instances of any size. We conduct a systematic empirical study across graph encoders, DRL algorithms, reward functions, training distributions, and hyperparameters. Evaluated zero-shot on the full BPPLIB benchmark against a constructive heuristic, a grouping genetic algorithm, and recent learned methods, our data-driven policy lowers the mean optimality gap of the constructive heuristic from 2.66% to 2.31%, with the largest gains on structured instances. Against learned baselines evaluated on the same benchmark, it attains a lower gap on most of the nine families and is far more stable across instance distributions. On the hardest benchmark family, it outperforms a state-of-the-art learned solver that relies on column generation and integer programming, while using no solver at all. A grouping genetic algorithm remains ahead overall, and we analyze where and why the residual gap arises.

Parallel coding agents can produce patches that work alone but fail when merged. This happens when one agent changes an interface or rule that another agent still relies on. We study these failures with stale, a benchmark for semantic coordination. Our evaluation runs the same tests on each patch alone and on their combination, counting only failures introduced by combining the patches. We use three tiers: synthetic tasks with controlled interface changes, pairs of merged pull requests, and constructed tasks that use real Django helpers. Among 834 runs on 417 mined Django pairs, only one showed interference after correcting the grading procedure. On constructed tasks using 12 Django helpers, interference occurred in 97% of runs. A message describing the completed concurrent change recovered 82% of runs. Reviewed pull requests may contain few unresolved parallel changes, even when agents fail on controlled tasks using real code. The constructed failure rates do not estimate how often these problems occur in practice.

It is well-known that any Coxeter group $W$ is a meet-semilattice with respect to weak order. Furthermore, when $W$ is finite, it is a lattice. In the first part of this paper, we give a recursive algorithm that computes joins in weak order for arbitrary finite $W$. In the second part, we complete the proof of a conjecture of Dyer expressing the inversion set of a join in terms of the union of inversion sets of each factor, under the assumption that $W$ is finite. Some proofs were developed in collaboration with Claude Opus 5.

We consider the problem of reconstructing a kernel $k$ and image $u$ in a blind deconvolution problem from convolutional data $g = k*u$. We particularly focus on the 3D case, i.e., on volumetric images. The problem is approached by imposing a semiparametric CP decomposition for the kernel and a variational TV penalty for the image. We also discuss a motion blur ambiguity effect, i.e., a nonuniqueness issue of the kernel in the blind deconvolution problem that may appear in certain videos of moving objects. The blind deconvolution algorithm uses an alternating minimization approach with TV regularization for $u$ and a CP decomposition step for the kernel. These are implemented with positivity projection, kernel normalization, and\textbackslash or causal support projection steps. Several numerical examples for 3D MRI images, hyperspectral images, and a grayscale test video show the effectiveness of the method.