国家预印本平台

This paper studies convolution rank regression (CRR) over decentralized distributed learning networks. We propose a novel decentralized CRR framework, in which estimators are obtained by solving consensus-constrained optimization with kernel-smoothed rank loss. The developed estimation scheme relies solely on local node data and information shared by neighboring nodes, thereby achieving privacy preservation and high communication efficiency. For heterogeneous network settings, we establish finite-sample error bounds for the decentralized CRR estimator and derive exact support recovery guarantees for the sparse decentralized CRR Lasso estimator. To facilitate numerical implementation, we adopt a generalized consensus ADMM to efficiently solve local subproblems across all network nodes. We verify the favorable performance of our developed approach via extensive numerical simulations and real-data experiments.

Few-shot multimodal classification commonly attaches a lightweight head, such as $k$-nearest neighbors, logistic regression, or a linear SVM, to a frozen pretrained encoder. Although computationally efficient, these heads can produce poorly calibrated confidence scores. We ask whether TabPFN can provide reliable confidence estimates on multimodal embeddings without sacrificing predictive accuracy, and under what conditions. We systematically evaluate TabPFN as a zero-gradient head for frozen image, text, and audio encoders. Across 22{,}820 evaluation episodes spanning 14 datasets, 11 encoders, and three modalities, TabPFN achieves the best mean rank among nine classification heads on both negative log-likelihood (NLL) and expected calibration error (ECE). At a representative setting, it reduces NLL by 48--62\% and ECE by 2.1--5.3$\times$ relative to the average of eight baselines while matching or exceeding their average accuracy. This calibration benefit transfers broadly, whereas the accuracy advantage is conditional: it concentrates at moderate-to-high shot counts and low-to-moderate feature dimensions ($k \ge 50$, $d \le 32$), and diminishes when labeled data are scarce, feature dimensions are high, or competing methods approach ceiling accuracy. After backbone adaptation, replacing the trained linear head with TabPFN improves calibration while preserving competitive accuracy, showing that representation adaptation and reliable head choice are complementary. Together, these results identify when TabPFN can serve as a training-free head for calibration-sensitive multimodal classification. To support transparency and reproducibility, we publicly release the source code, experiment configurations, and evaluation scripts in our GitHub repository: https://github.com/Jingxiang-Zhang/tabpfn-multimodal-embeddings.

We report a 15-day paper-traded deployment of an autonomous memecoin trading system on Solana decentralised exchanges (DEXs) as a controlled measurement of three microstructure questions on which classical equity theory offers well-defined predictions but on which the AMM-based Solana venue lacks published empirical measurement: time-of-day patterns on a 24/7 permissionless venue; whether decision-time filter stacks are net-positive against the counterfactual returns of the tokens they reject; and whether small-sample cumulative-return statistics on a heavy-tailed venue are structurally robust or fragile. The 190-trade sample (March 29 to April 12, 2026) shows a 40.5 percent win rate, mean per-trade return +0.62 percent, cumulative +117.7 percent (net SOL +0.03873), population skewness -1.21, excess kurtosis 6.61. Stratifying on the derived UTC entry hour (date minus heldMin), a Mann-Whitney U test of three exploratorily identified poorest-performing hours (2, 13, 23; combined n=17, mean -11.60 percent) against all other hours (n=173, mean +1.82 percent) yields U = 1,345, p = 0.5634 (two-sided asymptotic, continuity-corrected), rank-biserial r = 0.085; directional and non-confirmatory. A parallel counterfactual rejection-tracking system collected 4,874 forward-sample observations across 184 rejection events; 17.9 percent of events reached a 50 percent drawdown from reference within 24 hours. Removing the top three trades (1.6 percent of sample) flips cumulative return unprofitable. Alongside the trade log and rejection-sample corpus we deposit audit.py, an assertion-based reproduction script that exits zero iff every headline number reproduces from the deposited CSVs. Companion dataset: Zenodo concept DOI 10.5281/zenodo.20043301. The paper's principal contribution is measurement infrastructure and three transferable design lessons.

Translating user intent into physical radio signals is the last critical step in wireless prototyping. It chains protocol planning, baseband synthesis, and hardware configuration. Large language models and multi-agent systems have reshaped software engineering, raising the question of whether they can solve this problem. Yet current models fail at this task, even when augmented with domain tools. Because the stages run sequentially, an error at any stage propagates downstream, so the end-to-end success rate collapses toward zero even when each stage looks locally competent. We introduce RadioMaster, a fully autonomous multi-agent framework that drives user input to verified emissions transmitted over the air. It rests on three synergistic pillars. RadioWiki grounds generation in domain knowledge to suppress hallucination. RadioAgent decomposes the fragile pipeline into independently executable and locally recoverable stages. RadioEmulator gates deployment behind closed-loop physical-layer verification. We further build RadioBench, the first benchmark for autonomous radio signal generation. Extensive real-world evaluations show that RadioMaster substantially outperforms state-of-the-art baselines in configuration viability and signal fidelity, while reducing configuration time by up to 28x.

Vision-language models (VLMs) are increasingly applied to medical visual question answering (Med-VQA), yet whether they can \emph{localize} the evidence behind their answers---a prerequisite for clinical auditability---is poorly characterized. We separately evaluate VQA reasoning and visual grounding for four recent frontier VLMs (GPT-5.1, GPT-5.5, Gemini-2.5-Pro, Gemini-3-Flash), two domain-specific medical VLMs (Lingshu, MedGemma), and a dedicated open-vocabulary detector (Grounding DINO) on VQA-RAD and SLAKE. Two findings challenge the intuition that ``add grounding to improve VQA.'' First, \textbf{no model localizes medical targets well}: every off-the-shelf system---frontier, medical-specialized, or dedicated detector---scores mean IoU 0.05--0.24 on our SLAKE grounding split, at or barely above a trivial center-box baseline (0.10), with Acc@0.5 below 20\%. Second, and counter to the common ``localize-then-answer'' paradigm, \textbf{cropping to a bounding box degrades VQA even when the box is a perfect oracle}: on the matched subset where oracle ground-truth boxes are applied, GT-grounding \emph{lowers} closed-ended accuracy for every model (by 0.9--18.0 points versus using the full image)---consistent with the crop discarding global context the model relies on. Because the oracle box removes localization error by construction, the problem is not that perception is a recoverable bottleneck, but that grounding-by-cropping is itself the wrong interface. Finally, we show constructively that the two channels need not conflict: supervised fine-tuning of Qwen-2.5-VL-7B on answers \emph{alone} silently destroys box-evidence emission (0/418 parseable boxes), whereas mixing in a small amount of grounding supervision restores localization to 0.36 IoU---above every zero-shot model---while preserving answer accuracy.