Papers

Model failover restores availability, but changes which financial institutions share decision errors. We formulate outage-contingent routing through a local market-impact response matrix and study expected squared price displacement. A symmetric construction shows that a shared backup can leave an order-one concentration floor as the number of primary endpoints grows, while balanced fallback risk decreases inversely with the surviving endpoint count. For indivisible decisions, we derive the exact second moment of independent randomized routing and an effective-exposure granularity that determines its gap from fractional allocation. Conditional-expectation rounding gives a finite-agent bound without coupled quotas; a separate swap procedure preserves endpoint counts and is assessed against dual lower bounds. Across 60 synthetic portfolio networks and 11,340 scenario evaluations, the latter reduces risk by 6.57% and 10.53% for single and double endpoint removals at the central feedback setting with independent errors. A replay of 1,024 recorded API responses on constructed rebalancing tasks gives a smaller held-out reduction of 3.30% (paired bootstrap interval 2.07–4.57%). Strongly aligned errors, inferior endpoints, and indivisibility limit diversification. The contribution is an auditable routing stress test and implementation analysis, not an estimate of real-market crash probabilities.

Long-term memory agents increasingly rely on it- erative search and reusable experience to answer questions over large personal, factual, or narrative histories. However, current experience-memory systems largely optimize relevance: they re- trieve past search lessons that appear similar to the current state and inject them into the prompt. A relevant experience can still be harmful when the memory substrate, question intent, answer granularity, or evidence boundary changes. We propose CAVE- Mem, a training-free framework that represents experience as a typed intervention operator with applicability, boundary, and utility conditions. CAVE-Mem first obtains a base memory-search answer, then allows an operator to change it only if the oper- ator matches the current substrate, answer contract, evidence boundary, and cross-fitted utility; otherwise the system abstains. Experiments across long-term conversational memory, multi-hop question answering, and long-document narrative reasoning show consistent gains over relevance-only experience reuse.

Spacecraft thermal design requires repeated evaluation of how variations in the number and spatial arrangement of heat-generating components and in thermal boundary conditions affect the temperature field. High-fidelity numerical simulations are computationally expensive and therefore difficult to use for large-scale design screening. Although surrogate models can accelerate temperature-field prediction, existing approaches generally encode each complete configuration as a whole and do not explicitly exploit the reusability of local physical constituents across configurations, which limits their accuracy for component counts and combinations not covered during training. To address this issue, we propose the Tree-Structured Factor Composition Network (TFCN), which decomposes complex spacecraft thermal configurations into reusable local physical factors and employs a tree-structured composition module to learn the global temperature-field response associated with different factor combinations. TFCN is evaluated on two-dimensional steady-state spacecraft thermal-analysis cases with prescribed-temperature and radiative-flux boundary conditions. The model is trained exclusively on configurations containing no more than 15 heat-generating components and evaluated on unseen configurations containing 16-25 components. For the prescribed-temperature and radiative-flux cases, TFCN achieves component-count out-of-distribution RMSE values of 4.21 K and 18.62 K, respectively, representing reductions of 65.6% and 33.6% relative to the strongest baseline. These results demonstrate that TFCN improves the reliability of temperature-field prediction under variations in component count and provides an efficient surrogate for rapid spacecraft thermal-design evaluation and large-scale configuration screening.

Few-shot anomaly detectors are judged by ranking metrics, yet deployment requires an alarm threshold with a controlled false-alarm rate (FAR). We ask how much normal evidence, in images or category units, is needed to certify such a threshold for an unseen category. Using a frozen DINOv2 principal component analysis (PCA) residual ranker on 15 MVTec and 12 VisA categories under four corruption types, we show that target-only leave-one-image-out (LOIO) calibration is resolution-limited and shift-fragile: rank values cannot fall below $1/(k+1)$, and at the attainable level $\alpha=0.20$, empirical FAR reaches 0.341 on Gaussian-corrupted MVTec at $k=4$, 1.7 times the nominal level. A category-count feasibility calculus is then derived: even with all-zero category losses and no multiplicity charged, any deterministic, uniformly valid, distribution-free 95% upper confidence bound (UCB) requires at least 14, 29, and 59 independent and identically distributed (iid) category draws at $\alpha=0.20$, $0.10$, and $0.05$; these counts are necessary but not sufficient. The Cross-category Reliability Estimation with Source Support (CRESS) protocol splits source categories into disjoint reference, proposal, and certification roles. With only three or four certification categories, all 960 frozen configurations return the fail-closed threshold $\tau^\star=0$, and the smallest category-level UCB is 0.950. Image-unit analyses of the same archive select positive thresholds in 36.7% to 60.3% of target cells; these bounds hold for the selected source mixture, not for the marginal risk of a new-category draw. The contribution is a quantitative feasibility boundary and an estimand-aware protocol specifying when source evidence can, and cannot, support a transferable reliability claim.

We show that a three-body holonomy can select the conventional $SU(3)\to SU(2)\times U(1)$ breaking direction. Three particles are special in the present construction: after removal of the center-of-mass coordinate they are the minimal one-dimensional few-body system with a two-dimensional relative space, allowing a closed cycle around the triple-coincidence point once pairwise contacts are encoded by matching conditions. A cyclic three-state mass operator gives \[ M^{[k]}=M_0+2\Delta\cos\!\left(\frac{\theta_3+2\pi k}{3}\right),\qquad k=0,1,2. \] At $\theta_3=0$ the spectrum splits as $3\to2+1$, and the induced mass operator is proportional to the Cartan generator $T_8$. Its commutant is therefore $SU(2)\times U(1)$. In a trinification-like embedding, distinct left- and right-sector holonomies leave the hypercharge combination which reproduces the Standard-Model hypercharges for the usual trinification matter assignment. The construction is a toy mechanism for selecting the breaking direction; additional dynamics are required for rank reduction and vacuum selection.

"Train a model on the same problems written under two incompatible conventions, both correct, and ask what the ordering of that data writes into the parameters. The learning-rate schedule is not a background condition for that question. It is the averaging operator, and it decides the answer. We prove a bound in which the arrangement and the schedule enter the ordering effect as separate multiplied factors: the arrangement only as a block period, the schedule only as how much weight the endpoint can place on any one moment of the run. A decaying schedule cannot put a large step size and an uncontracted remainder at the same moment; a constant one does exactly that at the last step. That decay moderates ordering effects has been reported in pretraining; the mechanism, the separation, and a controlled measurement of both halves are ours. Ten orderings of one corpus, one budget, everything but the path held fixed, run twice under families differing in lr_scheduler_type and nothing else: at a constant rate the interior spans 0.2221 in allocation, 11.63 contrast floors, monotone in how blocked the arrangement is. Under the single cosine every published arm uses, the same ten arms occupy two distinguishable states where their own resolution would allow about ten. "Order matters" and "order does not matter" are the two ends of one knob. What the path writes is which convention the model commits to, and no exact-match benchmark can see it. Across twelve arms acc_A+acc_B is constant to within 9.7% while the allocation share runs 0.04 to 0.87, so the 12.29-sigma arrangement switch this paper measures is exactly zero under a convention-agnostic metric. That conservation is quoted from the decayed family throughout, the constant-rate one being a noisier place to read it. Marking the convention in the prompt collapses the switch and reaches 87.5% of the union ceiling."

When an informed adversary shares the audience of a constrained signalling channel, the signal that best protects the truth is the signal that best describes it. On 108 confirmatory items, the adversary-robust optimum aligns exactly with the salience pole from prior work. Across a 200,000-item pool, the two differ on only 2,748 items — lying exactly where the prior salience-to-Bayes coordinate is undefined. Where defined, robustness is achieved by moving from Bayesian discrimination entirely to salience. We show this by introducing an adversary to a forced-choice task (abstracted from Deception: Murder in Hong Kong). The adversary knows the target, observes the signal, and argues for the strongest wrong answer using a persuasion budget, $\beta$. As $\beta$ grows, the optimal signal shifts from the posterior-maximizing option to the margin-maximizing one; at $\beta = 0$, the game reproduces the original oracle model with a listener temperature of $\tau = 1$. This effect is real: 18.2 percent of the pool has an optimum that shifts under a finite budget, and each item's critical budget is exact. This coincidence structurally limits empirical evaluation. Two adversary framings change the chosen option of seven language models on 30 to 77 of 108 items against an exact no-effect rate. Yet, no measurement can determine whether this movement is toward the adversary-aware optimum or toward salience, because the two options are identical. This is a structural limit, not a null result. The diagnostic check is cheap: before evaluating adversary-awareness, verify whether the robust target coincides with a heuristic target on the evaluation items.

Fixed-size recurrent memory limits storage growth during inference, but successful recall depends on the task and training. We study a small DeltaNet variant with fixed token-specific key biases, trained to remember 32 new key-value pairings per sequence. With 32 KiB of recurrent matrix state, it achieves 99.95% mean accuracy across three training seeds when choosing among the sequence's values. Recall remains near perfect when filler extends the pre-query context to 1,798 tokens without adding pairings. Zeroing the first memory block removes this recall. An exploratory 48-pair test remains near chance after one quarter of the primary training budget and does not locate a capacity limit. Parameter-matched vector and Transformer baselines remain near chance, including the Transformer after additional training searches. This unresolved baseline failure prevents a memory-efficiency comparison.

Let $b_1,\dots,b_d\ge2$ be fixed integers, and let $k$ be their multiplicative rank. We study representations $n=a+b_1^{u_1}\cdots b_d^{u_d}$, where $a$ belongs to a set $\mathcal{A}$ of positive integers and $u_1,\dots,u_d$ are positive integers, counting distinct tuples of exponents separately. Under density and correlation assumptions on $\mathcal{A}$, we obtain a lower bound for the number of integers with many such representations, in terms of $d$ and $k$. In particular, when $\mathcal{A}$ is the set of primes or the set of positive integers representable as a sum of two squares, a positive proportion of the integers $n\le x$ have at least $c_1(\log x)^{d-1}$ or $c_1(\log x)^{d-1/2}$ representations, respectively, for some $c_1>0$ and all sufficiently large $x$. No multiplicative independence or coprimality assumptions on the bases are required.

We study the Kadison–Kastler distance and the interior angle between masas in finite-dimensional matrix algebras. In \(\mathbb{M}_2(\mathbb{C})\), we obtain an explicit formula for the Kadison–Kastler distance between two masas and show that every value in \([0,1]\) is attained by \(\mathrm{d}_{\textrm{KK}}(\Delta,u\Delta u^*)\). We further prove that the distance attains its maximal value precisely when the relative unitary is a Hadamard unitary. In addition, for every $v$ in the groupoid normaliser of $\Delta$, the Kadison–Kastler distance between \(\Delta\) and \(v\Delta v^*\) assumes either its minimal or maximal possible value. We also establish the identity \[ \mathrm{d}_{\textrm{KK}}(\mathcal{A},\mathcal{B})=\sin\alpha(\mathcal{A},\mathcal{B}) \] for any two intermediate subalgebras \(\mathbb{C}\subsetneq\mathcal{A},\mathcal{B}\subsetneq\mathbb{M}_2(\mathbb{C})\). Most notably, for any unitaries \(u,v\in\mathbb{M}_2(\mathbb{C})\), we establish the equivalence of maximal Kadison–Kastler distance, maximal interior angle, maximal Connes–Størmer modified entropy, and minimal Popa probabilistic index: \[ \mathrm{d}_{\textrm{KK}}(u\Delta u^*,v\Delta v^*)=1 \iff h(u\Delta u^*,v\Delta v^*)=\log 2 \iff \alpha(u\Delta u^*,v\Delta v^*)=\frac{\pi}{2} \iff \lambda(u\Delta u^*,v\Delta v^*)=\frac{1}{2}. \] For general \(\mathbb{M}_n(\mathbb{C})\), we prove a unitary invariance property for the interior angle and derive an explicit formula for the angle between arbitrary masas of the form \(u\Delta^{(n)}u^*\) and \(v\Delta^{(n)}v^*\). As a consequence, the interior angle is \(\frac{\pi}{2}\) precisely when the relative unitary \(u^*v\) is a Hadamard unitary. Finally, we show that every value in \(\left[0,\frac{\pi}{2}\right]\) is attained as the interior angle between a pair of masas in \(\mathbb{M}_n(\mathbb{C})\).

System prompts are the primary lever practitioners use to control language model behavior, yet what they actually do to the computation inside the transformer remains poorly understood. Across 17 instruction-tuned models spanning 8 architecture families and 1.5B to 72B parameters, we use Centered Kernel Alignment (CKA) to compare layer-wise representations under 20 system prompts in five functional categories. Effects are layer-selective and instruction-type-dependent: persona and formatting instructions deeply restructure intermediate representations, while safety instructions barely move them, producing changes statistically indistinguishable from a minimal baseline. Restrictive safety instructions and explicitly permissive ones ("you have no restrictions") engage near-identical computational pathways (mean CKA correlation 0.997), and this persists at commercial scale, where safety penetration remains below 10% even at 70B-72B. A linear probing baseline exposes the mechanism: the model encodes prompt category at every layer but restructures its computation only at a small subset, so the prompt is reliably "seen" but, for safety, not deeply "acted upon." Causal activation patching confirms these layers mediate behavioral change, and representational depth predicts behavioral effect size across the full 17-model cohort (Spearman rho = 0.761, p < 0.001). The findings provide a mechanistic explanation for the persistent jailbreak vulnerability of system-prompt-based safety. Code: https://github.com/Usama1002/system-prompt-illusion-cka

Intrinsic self-correction asks a language model to revise its own answer without receiving new external evidence. A second pass can recover mistakes, but it can also overturn answers that were already correct. We study this trade-off across 29 open-weight LLMs on BoolQ, GSM8K, and Corr2Cause by tracking correctness transitions between initial and revised answers. Aggregate accuracy can conceal substantially different revision behavior: for example, Llama-3.1-8B improves by 25.5 percentage points on GSM8K, while refinement changes 19.1% of initially correct answers into wrong ones. A controlled BoolQ study further shows that refinement prompts shift the balance between recovery and harm. We then compare three runtime choices: keeping the initial answer, always accepting the revision, and selectively invoking revision using signals available after the initial response. The comparison identifies settings where learned gating is useful and others where a simpler unconditional policy performs better. These results suggest treating intrinsic self-correction as a revision policy rather than as a uniformly beneficial second pass, and evaluating it through both the corrections it recovers and the errors it introduces.

Modern large language models now support context windows of more than one million tokens, which has raised the question of whether retrieval-augmented generation (RAG) is still necessary. Pure long-context (LC) processing is expensive and is known to under-attend to information placed in the middle of long inputs, while pure RAG is fast but bounded by retrieval quality and prone to errors when retrieved chunks are partially relevant or contradictory. We propose the Entropy-Driven Adaptive Router (EDAR), a framework that decides at inference time whether to answer a query from retrieved chunks or to escalate it to full long-context processing. The decision uses the predictive entropy of the token-level probability distribution computed over the first few generated tokens of the RAG response. The entropy threshold is selected on a held-out validation set by sweeping cost against accuracy. Experiments compare EDAR against pure-RAG and pure-LC baselines on LongBench v2 and Infinity-Bench. Predictive entropy correlates strongly with hallucination rate on a held-out set of 2,000 generations (Pearson r = 0.85, 95% CI [0.83, 0.87]). On the long-context benchmarks, EDAR retains 97.4% of the accuracy of the pure long-context baseline while reducing total token expenditure by 70.7%, escalating only 18.2% of incoming queries. The accuracy gap between EDAR and the pure long-context system is not statistically distinguishable from zero at standard sample sizes. Predictive entropy is a useful model-internal signal for routing between RAG and long-context inference, and a threshold-based hybrid system can recover most of the accuracy of long-context models at a small fraction of the cost. The framework does not depend on a specific retriever or LC backbone, and it does not require additional supervision beyond what is normally produced during decoding.

The Starterkit Lessons online and Starterkit Workshops held in Geneva each year have been the main method of onboarding newcomers to the LHCb experiment since its founding in 2015. The new software corresponding to Upgrade 1 of the LHCb and Run 3 of datataking at the LHC has necessitated a new version of this Starterkit to be written. This new version has made several improvements over the old one, with increased maintainability through testing of examples in CI pipelines, and has also allowed for translations of the Starterkit Lessons. In late 2025 the Run 3 Starterkit lessons were translated into Mandarin Chinese, to allow for a sibling Starterkit event to take place in China, with a dedicated liaison role set up to ensure synchronization between the two translations. To measure the success of the Chinese Starterkit Lessons, analytics have been added revealing that roughly 22% of all visits to the Starterkit website will access at least one Chinese-language page.

The Multi-Purpose Detector (MPD) experiment at NICA is designed to investigate strongly interacting matter at high net-baryon density through a broad program of heavy-ion measurements. Dielectrons constitute one of the key probes in this program, with contributions emitted throughout the entire space-time evolution of the collision. Their experimental measurement is, however, challenged by a large combinatorial background arising primarily from incompletely reconstructed photon conversions and Dalitz decays of light neutral mesons. In this work, we study the performance and capabilities of the MPD experiment for dielectron measurements using simulated minimum-bias $^{209}$Bi+$^{209}$Bi collisions at $\sqrt{s_{NN}}$ = 9.2 GeV. Electron identification is performed using the combined information from the TPC, the TOF detector and the ECal. A multilayer-perceptron classifier is used to optimize electron identification, leading to a substantial increase in electron detection efficiency relative to sequential one-dimensional selections while preserving an electron-sample purity close to unity over a broad momentum range. To address the dominant sources of combinatorial background, a pair-analysis strategy is developed that exploits partially reconstructed electron tracks. The method combines the pair opening angle, the TPC dE/dx signal, and an approximate reconstruction of the pair invariant mass to tag tracks from photon conversions and $\pi^0$ Dalitz decays at the pair level. In the mass interval 0.2 $< m_{ee}<0.7 $ GeV/$c^2$, the method improves the signal-to-background ratio by a factor of about 3.5 with the current track reconstruction algorithm. The study demonstrates the strong potential of the MPD experiment for dielectron measurements at NICA energies and highlights the importance of further improvements in low-$p_{\rm T}$ track reconstruction.

Mechanical metamaterials are materials that, by virtue of their microstructural architectures, exhibit mechanical properties not often found in nature: for example, auxetic mechanical metamaterials are materials that possess a negative Poisson's ratio in response to deforming forces. However, outside of relatively simple crystalline, lattice-based, or repeating-pattern architectures composed of mostly identical unit cells, mechanical metamaterials like auxetics are notoriously difficult to design, particularly across all three dimensions. Here we show that, for any structure that can be decomposed into triangular or tetrahedral meshes, those structures can be subjected to a simple transformation that that creates conditions within those triangular or tetrahedral simplexes that, regardless of their individual geometries, introduces a counter-rotating polygon/polyhedron mechanism that both guarantees a negative Poisson ratio and allows locally programmable mechanical properties. We apply this "shrink-and-wrap" construction to generate mechanical metamaterials from increasingly complex 3D structures, first from simple (convex) polyhedrons; then, to designs with arbitrarily large numbers of vertices and designed using constructive solid geometry (CSG) and 3D scanning; and, ultimately to 3D objects generated from videos captured by mobile phone camera – making potentially any object or geometric structure designed or found in the real world transformable into a mechanical metamaterial. We show that these monolithic auxetic constructions remain readily 3D-printable via additive manufacturing techniques. We expect this piecewise approach to designing arbitrarily complex mechanical metamaterial structures can enable numerous potential applications where programmable internal reconfigurations and/or force redirection are required across and throughout their 3D geometries.

Can a robot improve itself the way coding agents now improve software? We built an agentic system to find out. It watches a robot fail, works out which capability is missing, writes new skills or finds and installs external models, tests every change in simulation, and repeats, with no human writing robot code. We ran it for 123 improvement rounds on household manipulation tasks. This report describes what we learned. The good news is that the agent can discover capabilities on its own: noticing that its targets were out of view, it asked for an active-viewing model, debugged it, and deployed a working search skill. The bad news is that its improvements did not add up. Changes kept passing their tests, yet the target task, putting condiments on the top shelf of a fridge, never succeeded. We found that the agent was rarely the bottleneck. Three things around it were. First, chained perception modules do not understand relations. Segmenters such as SAM 3 find shelves but not "the top shelf", so the agent filled the gap with ever more geometric rules that never converged, when what it needed was a different kind of model. Second, skill chains lock learning onto the first step. Long tasks mostly fail early, so evidence and fixes pile up there, and later skills are rarely reached, tested, or improved. Third, what the agent learns is decided by the harness. The agent optimized exactly what the evaluator measured, including where it was wrong, and weak tests and misleading memory turned activity into a standstill. We distill these lessons into concrete recommendations for building robot systems that improve themselves, each paired with an experiment that could prove it wrong.

Domain Generalization (DG) aims to develop models that remain robust to conditions unseen during training. While DG has been systematically studied in computer vision through controlled benchmarks and diverse distribution shifts, its evaluation in spoken language recognition remains less structured. Existing speech datasets provide valuable benchmarks for robustness to real-world acoustic conditions, but are primarily designed around specific scenarios and large scale rather than as general-purpose tools for systematically constructing and evaluating domain shifts. In this work, we introduce a smallscale speech dataset and evaluation protocol for controlled studies of acoustic domain shifts. It enables the evaluation of spoken language recognition models under a variety of acousitc domain shifts. We introduce the speech modality into the DomainBed Domain Generalization framework and evaluate three domain generalization algorithms. We show that in-domain performance is not a reliable predictor of cross-domain robustness and verify that explicit domain invariant algorithms such as MMD or DANN algorithms do not outperform ERM. We further validate the generality of these findings on MMS-LID-126, a state-of-the-art spoken language identification system. We release the code.

Does high image fidelity imply readable screen content in 3D Gaussian Splatting (3DGS)? We introduce 3DGS-SC, a controlled static screen-content dataset and benchmark for examining this mismatch. Ten procedural scenes provide fixed multi-view splits, exact cameras, screen masks, text boxes, and transcripts. The protocol separates whole-image fidelity, screen-region fidelity, OCR readability, and edge preservation. In the reported five-method comparison, LightGaussian exceeds Mip-Splatting in screen PSNR by only 0.08 dB, yet trails it in OCR accuracy by 12.7 percentage points. Across all ten method pairs, screen-PSNR and OCR orderings disagree in seven cases. These aggregate results expose a method-selection failure of fidelity-only evaluation. Complementing prior text-aware 3DGS research, 3DGS-SC targets controlled monitor interfaces and exact annotations; scene-wise robustness and acquisition effects remain open validation questions.

Panoramic $360^\circ$ scene understanding increasingly relies on icosphere transformers, but a state-of-the-art spherical model loses more than half of its segmentation accuracy when the camera rotates by $90^\circ$, and controlled ablations identify gauge dependence in its relative-position bias as a major contributor. We propose gauge-equivariant relative position encoding (GE-RPE): a parameter-free Reynolds average of the bias over a finite cyclic subgroup $C_n\!\subset\!\mathrm{SO}(2)$ of gauge rotations. Plugged into a SphereUFormer backbone the change is invisible at deployment—zero added parameters and $1.5$–$4.4\%$ forward latency—and the matched three-seed GE-RPE model records a $1.3\%$ drop; the published SphereUFormer checkpoint records $53\%$ under the same stress protocol but a different training recipe. Once the gauge defect is removed and a teacher-token permutation $\pi_R$ aligns the SSL views to the rotated student frame, iBOT$+$MAE pretraining stops being a liability and becomes a clean low-label lever: the full framework EquiSSL (GE-RPE $+$ $\pi_R$ $+$ iBOT$+$MAE) tightens the drop to $0.8\%$ at $68.30\%$ val mIoU and lifts $1\%$-label fine-tuning by $+2.39$ mIoU on the $N{=}373$ test split (and by $+4.10$ on the smaller $N{=}40$ val split); the same fix carries over to monocular depth and to zero-shot Structured3D segmentation. The construction is provably $C_n$-invariant and $\mathcal{O}(n^{-2})$-close to the continuous $\mathrm{SO}(2)$ average, making the resulting model a usable $360^\circ$ visual-computing primitive across panoramic relighting, immersive video, and cross-dataset transfer. Code is available at https://github.com/Jaywalk18/equissl-release.

Vision-based tactile sensors are a popular solution for capturing rich contact surface geometry. However, to obtain high-detail contact surface normals and depth, it is necessary to calibrate the sensor by physically pressing a probe with known geometry against the sensor elastomer and mapping tactile images onto the ground truth probe's shape. This approach does not scale across different tactile sensors, and the calibration effort can be complex depending on the sensor shape and optical system. Instead, we propose a calibration-free procedure for the estimation of contact surface normals using Universal Photometric Stereo neural networks. In a series of real-world experiments, we evaluate our approach on 3 sensors with different optical systems, demonstrating that universal methods are a suitable approach for estimating surface normals at the contact patch from tactile images, thereby alleviating the need for tactile sensor calibration. Controlled experiments with a metal ball show that universal methods match the calibrated method, with a mean angular error of $6.56^{\Large\circ}$. We show that the proposed framework recovers high-frequency surface details of objects with natural textures, achieving an overall mean angular error of $10.66^{\Large\circ}$. Universal method robustly recovers the contact surface normals captured with dome-shaped Digit 360, achieving a low angular discrepancy of $10.18^{\Large\circ}$ relative to the calibrated baseline. This experiment demonstrates that with sufficient illumination settings surface normals could be estimated using a model trained solely on synthetic data. By providing a unified representation of contact surfaces across different vision-based tactile sensor designs, Universal Photometric Stereo neural networks lay the foundation for transferable tactile perception across sensors.

Video anomaly detection for patrol robots and surveillance systems must recognize abnormal interactions among familiar entities. Existing pixel-reconstruction and isolated-entity methods may fail when individual entities appear normal but their spatial or motion relations are abnormal. This work presents Interaction-Centric Network for Temporal Entity-Relation Analysis and Consistency Testing (INTERACT), a framework for unsupervised video anomaly detection. Person-Object Appearance-Motion Interaction (POAMI) moves beyond isolated-entity encoding by jointly modeling the appearance, motion, and spatial configurations of persons and objects in each frame. The resulting representations capture both individual entity states and their cross-entity context. Based on these representations, Target Geometry-Guided Relational Interaction Prediction (TGRIP) predicts target entity interaction states from historical relational memory and target-frame geometry without explicit identity tracking. Motion-Interaction Reconstruction and Alignment (MIRA) then evaluates these predictions through conditional flow reconstruction and semantic consistency checking, providing complementary anomaly evidence beyond prediction error alone. INTERACT achieves state-of-the-art performance, obtaining a frame-level AUC of 84.5% on the ShanghaiTech benchmark. Ablation studies show that removing cross-entity attention causes the largest performance drop, demonstrating the necessity of relational modeling. INTERACT is particularly effective for anomalies caused by changes in relations among people and objects, while maintaining competitive performance in general scenarios. The source code will be made publicly available at https://github.com/ppworkhard/INTERACT.

Access to past reports changes how consumers learn about competing information providers. A subscriber who has learned to interpret a familiar source may find an equally competent rival less useful. We study how opening the rival's archive changes renewal prices and the effect of outcome feedback. Better feedback strengthens the familiar source's advantage when rival reports are inaccessible, but can weaken it once some rival records are available. We derive the exact threshold in a Gaussian learning model with differentiated price competition. Opening further records narrows the renewal premium and raises consumer surplus. When the familiar source initially serves a sufficiently large share, however, partial opening can reduce aggregate forecasting quality and total surplus by drawing consumers toward a rival that remains less informative. Complete access improves both outcomes relative to every incomplete archive. These results concern a covered interior renewal market with inherited customer histories, before implementation costs. The results distinguish the quality of feedback from access to the records with which consumers combine it.

Vision-Language Models (VLMs) have advanced rapidly in static visual understanding, yet remain unreliable when judging how an egocentric task is progressing. Given a task instruction and two visual observations, a model should determine which state is closer to the goal by analyzing task-relevant object configurations and spatial relations, rather than relying on timestamps or presentation order. This distinction is critical in manipulation, where retries, corrective actions, and temporary regressions make progress inherently non-monotonic. We introduce EgoTSR, a unified framework for diagnosing and improving order-robust task-progress understanding. First, SpatialLogic-Bench evaluates each physical state pair in both original and order-swapped presentations across short- and long-horizon settings, exposing whether a model follows task-state evidence or chronological shortcuts. Second, our data construction pipeline converts successful, approximately monotonic manipulation and first-person trajectories into bidirectional supervision; LongTag further preserves intermediate subtask structure for long-horizon comparison, while failure-aware data extend learning to regressions and recoveries. Third, a progressive CoT-to-Tag curriculum first supervises evidence-grounded interpretation of task-relevant state changes and then consolidates the comparison rule through scalable label-only training. Experiments reveal substantial input-order bias in representative VLMs. EgoTSR achieves 92.4% long-horizon accuracy with a 0.1-point forward-inverse Gap. Failure-aware supervision further improves accuracy on non-monotonic trajectories by 11.8 points and Recovery Accuracy by 11.2 points, while maintaining broad visual and spatial capabilities. These results establish goal-conditioned state comparison as an explicit formulation of egocentric spatiotemporal reasoning for task-progress understanding.

Recognition of student affective states from classroom video is constrained by a class imbalance problem whose true nature is, we argue, under-analyzed. We show that on DAiSEE – the de facto benchmark for this task – class imbalance is fundamentally a dimension-specific phenomenon: raw imbalance ratios reach as high as approximately 79:1 (Frustration), and the structure of imbalance – not only its magnitude – varies across dimensions, so uniform algorithmic treatments (aggressive class reweighting, focal loss, uniform binary simplification) fail in dimension-dependent ways. We propose the Adaptive Hybrid Label Strategy (AHLS), which assigns four-level classification to dimensions with manageable imbalance and binary classification to severely long-tailed dimensions, coupled with a null-class placeholder mechanism that stabilizes multi-task optimization by compressing the maximum effective inverse-frequency weight ratio from as high as ~79$\times$ to at most ~8.5$\times$ across the four dimensions. Validated on a lightweight FERShuffleNetV2 + LTCN architecture (0.39 M parameters, 0.07 G FLOPs), the strategy attains the highest mean Macro-F1 of 47.70% across all four affective dimensions among 11 compared methods, including five state-of-the-art deep models (up to 167$\times$ larger) and five traditional machine-learning baselines. A cross-model behavioral analysis on DAiSEE indicates that, in this setting, classification strategy may influence minority-state recognition more strongly than further backbone scaling alone. We also report evidence of an annotation-density limitation of DAiSEE on minority affective states, which suggests that benchmark design may bound future progress as much as algorithmic refinement. Index Terms Affective computing, classroom video analysis, class imbalance, multi-task learning, lightweight deep learning, DAiSEE, engagement recognition.

We built an ordinary stacking ensemble for AI-generated image detection – three open detectors producing five scores, fused by a gradient-boosted meta-learner that treats a detector's failure as missing data – deployed it, and then evaluated it against three controls it should have faced first. This paper reports what the controls found, including where they overturned our own earlier conclusions. Fusion is worth its cost only when refitted on the target domain. The shipped meta-learner, fitted on a separate corpus, does not beat its best single member on 2000 StyleGAN faces (AUC 0.9896 vs 0.9961; McNemar p = 1.000). But a stacker refitted in-domain beats that member plus a post-hoc calibrator (Delta-AUC = +0.0025, [+0.0014, +0.0038]; p = 3.4e-4). An earlier draft claimed the calibrated single detector won outright; that comparison mixed regimes and we correct it here. One corpus is not an evaluation. On 80 screenshots every model's AUC interval contains 0.5. We can say nothing stronger: the difference between the ensemble's drop and its best member's is [-0.185, +0.115]. Abstention is real, correlated, and mishandled. With four of five detectors silent and the survivor reporting "real", the system returns P(AI) = 0.9985, because an all-NaN input scores 0.9995 in a learner never fitted with missingness. The three AIDE checkpoints fail together, sharing one preprocessing path. A quorum rule requiring two distinct architectures prevents both failures with no retraining. We also find Corpus A carries a class-conditional JPEG bias severe enough to separate the classes from the header alone, which limits every in-domain number we report. Code, harness, hash-identified artifacts and all corrections are released.

We examine the nonlinear initial value problem $\dot{y} = C(1 - y^\alpha)$ modeling polytropic gas permeation. For rational $\alpha$, we derive the exact primitive via cyclotomic partial fraction decomposition, establishing its real trigonometric representation. By synthesizing Ritt's theorem with Rosenlicht's differential algebraic residue criterion, we prove the exact step response $y(t)$ is an elementary function if and only if $\alpha \in \{1, 2\}$. Admitting the Lambert $W$ function extends closed-form solvability strictly to $\alpha \in \{-1, 1/2\}$. Finally, the asymptotic limit $\alpha \to 0$ converges to a singular dynamics solved by the logarithmic integral. These results constitute a complete classification of all explicit closed-form solutions admitted by this system.

Multimodal large language models (MLLMs) must balance local detail against scene context when interpreting ultra-high-resolution (UHR) remote sensing (RS) imagery within a limited visual-input budget. Existing selection-based methods either prune tokens and select patches through relevance scoring, or crop actively through repeated inspection. Neither strategy directly redistributes pixels within a continuous full-scene view: the first retains selected tokens or patches, and the second re-encodes a crop detached from its surroundings. Our pilot study finds that a frozen MLLM already produces useful question-guided spatial requests, yet crop-based inspection of the selected regions does not consistently improve its answers. We therefore formulate UHR understanding as a question of where to spend a fixed pixel budget. Based on this, we introduce GazeEarth, a simple-yet-effective training-free framework that couples question-guided region selection with full-scene foveated observation. The MLLM selects evidence cells from an indexed overview; a deterministic, topology-preserving warp resamples the original image onto a fixed-size canvas, enlarging their shared neighborhood while compressing the periphery; the same frozen model answers from this focused view, using at most two MLLM calls and no external selector or iterative search. Across three UHR remote sensing benchmarks and four frozen backbones, GazeEarth improves benchmark-averaged accuracy by 4.6 to 9.4 percentage points over direct answering and 3.4 to 4.3 over overview answering, outperforming task-trained methods. Our analyses show that existing MLLMs can guide where to look in UHR images on their own, and that what they can infer from the selected evidence depends on how that evidence is presented.

Sub-billion-parameter Vision-Language Models are increasingly viable for on-device deployment, yet compact model size alone does not guarantee usability. On a phone, such a model can still require more than two minutes to produce its first token. On-device usability depends on three axes: accuracy, efficiency, and behavioral reliability; standard benchmarks miss the third, with answers too short to expose doom loops and prompts too benign to probe adversarial safety. We introduce a diagnosis-driven post-training recipe in which a teacher VLM stress-tests the student, uncovers failure modes beyond human priors, and converts them into targeted supervision and preference alignment, supplementing generic data scaling with failure-driven optimization. Coupled with two visual-token policies, the recipe yields two accuracy-efficiency variants with improved behavioral reliability. \NanoFull attains a 62.3 normalized average over 17 benchmarks, the highest among openly released $\sim$0.5B models, +7.4 over its base at identical architecture and token budget, with doom-loop rates at or below the strongest baseline's. \FlashFull retains 61.4 while cutting warm time-to-first-token on a Pixel 9 from 138\,s to 6.1\,s (23$\times$). By jointly addressing all three axes, we move compact VLMs toward practical on-device usability.

We study how learning about a forecasting method affects a seller's choice of evidence. More evidence improves the current forecast but makes the method's persistent component harder to identify. A monopolist may therefore collect less evidence initially to help customers interpret later forecasts. Because the seller captures only part of customers' gains from costly evidence, it switches to less initial evidence sooner than a welfare-maximizing regulator. A minimum evidence requirement improves welfare only for intermediate weights on future benefits. An independent assessment removes this learning trade-off but may be privately unprofitable despite being socially worthwhile.