跳到正文
研踪PaperTrail

论文库

检索计算机领域论文，按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+0新增论文
+0其中 AI 相关
+0已有论文修订
+93新增 AI 导读

arXiv 最近完成扫描：2026-11-16 01:34 · 全站最近入库：2026-11-14 15:09各信源进度 →

统计口径与数据新鲜度

统计 2026-11-15 03:00 — 2026-11-16 03:00（北京时间） · 点击数字查看对应简报；AI 相关包含在新增论文内。

最近完成检查：2026-11-16 01:34最近发现新增版本：2026-11-14 15:09库内论文最新更新时间：2026-11-13 02:59

每分钟刷新已保存的数据。空检查不代表采集异常；新收录可能包含历史论文。AI 导读按新增版本计数。

23,348 已收录论文19,580 AI 相关6,729 有修订112 HF 社区发现

浏览论文库

研究日报 / 周报 →按信源浏览 →

时间范围跟随排序：版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常，查看状态 →

找到 23,348 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-16 01:33（北京时间）
新论文 · v1arXiv:2611.16793

EG360: Efficient and Generalizable $360^\circ$ Depth Estimation via Adaptive Panorama-Perspective Distillation and Spherical Cross-Level Feature Fusion

Laiyan Ding, Chongxi Meng, Rui Huang

原文摘要

Generalized panorama depth estimation has witnessed great advances in recent years. These methods typically achieve robust zero-shot generalization performance using large pretrained models, but their high computational cost makes real-time inference on embedded platforms challenging. In contrast, previous lightweight panorama models have mostly been developed for single-domain evaluations and cannot generalize well due to domain shifts. To reduce the gap between these two types of models, we introduce EG360, an efficient and generalizable panorama depth estimation model. Specifically, building on a lightweight depth model, we design an Adaptive Panorama-Perspective Distillation (APPD) strategy guided by scene geometry, along with a Spherical Cross-Level Feature Fusion (SCFF) module for global semantic perception. As a result, with only 6.593M parameters after reparameterization, EG360 achieves strong zero-shot performance on real-world benchmarks and runs at 35 FPS on a Jetson Orin NX 16GB in MAXN mode, whereas a previous ViT-small model runs at less than 2 FPS. Our code will also be made public for future research.

cs.CV

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16792

Systematic Evaluation of Training Data Generation Strategies for Machine Learning-Based Molecular and Materials Property Prediction

Minh Nghia Nguyen, Beate Paulus, Akhil S. Nair

原文摘要

Machine-learning (ML) models for molecular and materials property prediction are often trained on large datasets curated through arbitrary or generic sampling strategies. Such approaches can lead to the over- or under-representation of regions of chemical and materials space that are most relevant to the prediction task. An important challenge is therefore how to efficiently curate training data to maximize predictive performance while minimising the number of labelled examples required. Here, we systematically compare training data acquisition strategies for molecular and materials property prediction, ranging from random sampling to various active learning approaches. We evaluate their performance across different model classes, representations, datasets, and target properties. Our results show that the effectiveness of data acquisition strategies depend strongly on the interplay between the model, dataset, and prediction task, although uncertainty-based sampling generally outperforms alternative strategies. Based on these findings, we derive guidelines for selecting data acquisition strategies under a given labelling budget.

cond-mat.mtrl-sci · cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16789

ScrewSense: Mapping the Invisible - Anomaly Detection and Error Classification in Screwdriving Processes

Tobias Lettner, Tatyana Ivanovska, Anwar Al Assadi, Christian Bergler

原文摘要

The growing volume of process data generated by screwdriving systems, combined with advancements in computing power, data storage, and algorithms, is facilitating the application of artificial intelligence in this field. Accurately detecting and classifying screwdriving errors is crucial for ensuring product quality, avoiding costly rework in manufacturing, and supporting sustainability by facilitating the maintenance, reuse, and recycling of screw-connected products. In this context, machine learning approaches offer a promising alternative to traditional monitoring techniques, such as window-based methods or envelope curves. This study examines the use of deep learning techniques for anomaly detection and error classification in screwdriving processes. The data corpus consists of multivariate time series capturing process parameters, such as torque and rotation angle. For anomaly detection, the embedding space of a vector quantized variational autoencoder was analyzed. In terms of error classification, four different model architectures were evaluated. The results demonstrate the effectiveness of anomaly detection and confirm that deep learning models can distinguish between multiple error categories beyond traditional binary classification. This extends the capabilities of conventional monitoring approaches, achieving accuracies of up to 100%. Additionally, real-world data material was examined in order to further demonstrate model robustness. Overall the results indicate strong generalization capability and highlight the practical potential of the developed models for use in industrial screwdriving processes, eliminating the need for expert knowledge in error identification and significantly reducing manual effort and analysis time.

cs.LG · cs.RO · eess.SP

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.12038

DPGS-SLAM: Diffusion-Prior and Provenance-Aware Monocular Gaussian Splatting SLAM for Dynamic Visual Sensing

Yifan Luo, Takafumi Matsumaru, Vibekananda Dutta, Xin He

原文摘要

This paper presents DPGS-SLAM (Diffusion-Prior and Provenance-Aware Monocular Gaussian Splatting SLAM) for dense mapping in environments containing dynamic objects. Objects in dynamic motion remain a significant challenge for dense visual SLAM since dynamic observations can contaminate camera tracking and scene reconstruction. Although 3DGS (3D Gaussian splatting) enables efficient dense mapping, existing 3DGS-SLAM systems may exhibit instability when dynamic objects violate their underlying static-scene assumption. Excluding dynamic observations can reduce their influence on tracking and mapping, but may leave occluded static backgrounds missing or weakly constrained. Our proposed DPGS-SLAM is a monocular RGB SLAM framework that supports provisional representations of insufficiently observed static backgrounds through asynchronous diffusion-based background completion and provenance-aware Gaussian mapping. Dynamic regions are identified through optical-flow-based motion discovery and temporally consistent mask propagation by SAM2. For selected keyframes, a fine-tuned diffusion inpainting model generates plausible background appearance in occluded regions, while metric depth predicted from completed images initializes provisional Gaussian primitives. Provenance labels distinguish generated primitives from observation-supported primitives, excluding generated content from camera-pose estimation and observed-map optimization. When previously occluded backgrounds become visible, eligible generated primitives are removed, and observation-supported primitives are independently initialized from real RGB observations and predicted depth. Experimental results demonstrated competitive camera tracking, with average ATE RMSE values of 1.2 cm on TUM RGB-D and 2.7 cm on BONN RGB-D datasets. On StaDy4D, DPGS-SLAM improved average PSNR by 4.60 dB over WildGS-SLAM using only monocular RGB input.

cs.RO · cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v4arXiv:2611.05355

Reading Why: Born-Measured Attention with Exact Word Evidence on Light and Qubits

Teerawat Tongloy, Jiraporn Pongsopa, Rangsan Jomtarak

原文摘要

When a language model answers, we can usually only estimate why. We build a transformer in which every nonlinearity is a quadratic detection - an interference, a balanced reading or a camera - and the output is a Born probability: attention |<q|k>|^2, the form of the Born rule, normalised per row, a measured product for the feed-forward nonlinearity, and a Born readout; between detections, classical control rescales and re-encodes. With its denominators held at the input, each class probability is a polynomial in the word amplitudes, so its integrated-gradient evidence is computed by a finite quadrature with no quadrature error: the words' contributions sum to the class probability above its no-information share to 10^-13. On e-SNLI natural language inference it reaches 77.45 +- 0.91 % accuracy, level with a same-size softmax transformer (76.62 +- 0.36 %) and an order-free bag of embeddings (77.79 %). Exactness buys the sum, not the best ranking. Deleting the word its evidence ranks first changes the answer in 44 % of pairs, of the 53 % some deletion changes, against 31-32 % for the softmax model's integrated gradients; the model's own gradient x input does as well (43 %), as does a bag of embeddings' (43 %, seed 0), and leave-one-out does better. Holding the denominators helps the low end: the five words ranked last change the answer in 1 %, against 6 % free. The same weights compile exactly to simulated optoelectronics, and their attention to qubits. On ibm_kingston, reading only the attention, the device gave the model's answers in all three Hadamard-test runs (the two shortest pairs and one without its deciding word) and both compute-uncompute runs (one circuit per overlap); uniform attention answers otherwise on two of the three inputs. With the device's row sums the evidence correlated with the exact one at 0.997 and 0.962 by Hadamard tests; with compute-uncompute, 0.971 and 0.48.

quant-ph · cs.CL · cs.LG · physics.optics

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16787

MESH: Multi-receiver Evaluation of Spatial Hearing across Egocentric and Allocentric Views

Wen Wen, Bohan Li, KwanHo Tam, Yu Xi, Jing Peng, Shuai Wang, Kai Yu

原文摘要

Spatial hearing in multi-receiver environments requires identifying who speaks, where they are, and how they move. Existing benchmarks evaluate spatial prediction, transcription, tracking, and distributed speech processing, but offer limited evaluation of whether speech signal remains associated with the correct physical source across receivers and motion. We introduce MESH, which evaluates this association through egocentric and allocentric views of shared acoustic scenes. The egocentric view uses local recordings and peer metadata, while the allocentric view combines synchronized recordings from multiple receivers in a shared 3D world frame. MESH contains nearly 40,000 questions across 13 static and dynamic localization and spatial transcription tasks. We evaluate eleven audio-model configurations and a language-model agent using specialist tools. The best static distance-estimation and dynamic direction-tracking accuracies are only 21.90% and 41.54%, respectively, while the lowest word error rates across four transcription tasks range from 59.15% to 73.10%. Ablation studies and additional experiments reveal that existing models are not yet able to effectively leverage peer information and struggle to handle multi-speaker and dynamic scenarios. These findings highlight the gap between access to distributed acoustic information and its effective use for spatial speech understanding. We release the dataset and evaluation code to support further research on collaborative and distributed spatial hearing.

eess.AS · cs.SD

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2509.03244

FoMEMO: Towards Foundation Models for Expensive Multi-objective Optimization

Yiming Yao, Fei Liu, Liang Zhao, Xi Lin, Yilu Liu, Qingfu Zhang

原文摘要

Expensive multi-objective optimization is a prevalent and crucial concern in many real-world scenarios, where sample-efficiency is vital due to the limited evaluations to recover the true Pareto front for decision making. Existing works either involve repeatedly fitting Gaussian process surrogates from scratch for each newly encountered problem, or rely on computationally demanding hypervolume-oriented policy learning for amortized optimization, making it challenging to achieve scalable pre-training and efficient yet robust adaptation for diverse emerging real-world applications. To address these challenges, we propose FoMEMO (Foundation Models for Expensive Multi-objective Optimization), which adopts a decomposition-based training paradigm that converts multi-objective optimization into preference-wise aggregated posterior learning tasks, enabling scalable pre-training on hundreds of millions of diverse synthetic datasets without relying on extensive real-world domain experiments. At test time, given observed trajectories from unseen problems and user preferences, the foundation model performs in-context posterior prediction and optimizes the derived acquisition functions for efficient candidate generation, achieving strong generalization and optimization performance without any subsequent model training or updates.

cs.LG · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2512.15033

Beyond Accuracy: A Geometric Stability Analysis of Large Language Models in Chess Evaluation

Qingya Hu, Ruifeng Cao, Weiqi Wang, Xidan Song

原文摘要

Large language models (LLMs) are usually evaluated in chess by how closely their position evaluations agree with a strong engine, but such agreement does not show that a model evaluates equivalent positions consistently. We apply metamorphic relations to chess evaluation: a 180-degree rotation of pawnless positions, left-right mirroring, color swapping, position pairs one move apart, conversion between Forsyth-Edwards Notation and a natural-language description, and illegal positions. Six LLMs (Claude Sonnet 4.5, DeepSeek Chat, Gemini 2.5 Flash, GPT-5.1, Grok 4.1 Fast and Kimi K2 Turbo) evaluated seven test sets of 500 items each, with Stockfish 16.1 as the reference. All six models failed on rotation: at most 10.6% of the 500 rotated pairs received evaluations within 300 centipawns (cp) of each other, and the models often declared the legal rotated positions illegal. On the other four transformations, stability ranked the models as accuracy did except for the first two places: GPT-5.1 had the lowest mean stability error (186.17 cp) and Claude Sonnet 4.5 the lowest accuracy error (330.41 cp). Rejection rates for illegal positions ranged from 44.2% to 96.0%. Stability should be reported with the share of consistent pairs, because means over answered pairs concealed most rotation failures.

cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2509.11789

costream: A Streaming Evaluation Protocol for Wearable Fall Detection

Timilehin B. Aderinola, Ilaria D'Ascanio, Luca Palmerini, Lorenzo Chiari, Jochen Klenk, Clemens Becker, Brian Caulfield, Georgiana Ifrim

原文摘要

Wearable fall detectors are commonly evaluated on presegmented windows, where a fall window is extracted at an annotated impact and non-fall windows are sampled from fall-free signal. However, a deployed detector processes a continuous sensor stream in which the location of a fall is unknown. We specify a streaming evaluation protocol in which a detector traverses an unsegmented recording, aggregates window-level scores into a temporal confidence map, and emits timestamped events that are evaluated against the annotated fall. Holding the dataset, participant-wise splits, window construction and fitted models fixed, we compare presegmented and streaming evaluation on clinically verified real-world falls. The two protocols identify the same best-performing method overall, but select different methods in 6 of 20 model-window configurations, with the divergence associated with differences in the temporal clustering of false alarms. Streaming evaluation also exposes deployment-relevant quantities, including false alarm rate over elapsed monitoring time, detection delay, and the effect of the decision threshold, that are not captured by per-window evaluation. These results show that evaluation protocol can materially affect conclusions about wearable fall detectors and should be considered when assessing their suitability for continuous monitoring.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2601.14175

A model of errors in transformers

Suvrat Raju, Praneeth Netrapalli

原文摘要

We study the error rate of LLMs on tasks like arithmetic that require a deterministic output, and repetitive processing of tokens drawn from a small set of alternatives. We argue that incorrect predictions arise when small errors in the attention mechanism accumulate to cross a threshold, and use this insight to derive a quantitative two-parameter relationship between the accuracy and the complexity of the task. The two parameters vary with the prompt and the model; they can be interpreted in terms of an elementary noise rate, and the number of plausible erroneous tokens that can be predicted. Our analysis is inspired by an ``effective field theory'' perspective: the LLM's many raw parameters can be reorganized into just two parameters that govern the error rate. We perform extensive empirical tests, using Gemini 2.5 Flash, Gemini 2.5 Pro and DeepSeek R1, and find excellent agreement between the predicted and observed accuracy for a variety of tasks, although we also identify deviations in some cases. Our model provides an alternative to suggestions that errors made by LLMs on long repetitive tasks indicate the ``collapse of reasoning'', or an inability to express ``compositional'' functions. Finally, we show how to construct prompts to reduce the error rate.

cs.LG · cs.AI · cs.CL · hep-th

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16779

Computation with Backward Feedback: Power, Observation, and Cost

Saman Farhat, Farzad Didehvar

原文摘要

How much can a computation gain by passing information back to an earlier stage, and how soon can an observer obtain the answer? We study deterministic programs whose runs are accessed at randomly selected stages. A program keeps a register, a finite block of memory. A random clock selects the current stage; each tick takes a random step, with a law fixed independently of input, program and register. When the clock moves backward, the register contents at the stage it leaves start a new run, and the clock continues where it lands. Equilibrium readout requires a distribution of clock position and register that one tick leaves unchanged, and a correct final answer under every such distribution. For bounded steps in both directions with negative average, we obtain an exact classification at each polynomial scale s. If running time and register length are both O(s), feedback gives exactly nondeterministic space O(s); without transmission the class is deterministic time O(s), which is strictly smaller. An observer that starts at stage zero with an all-zero register, reads sequentially and never accepts a nonmember sees, with fewer ticks than any exponential, only languages fixed by a bounded input prefix. Exponentially many ticks reveal every linear-time language. For a clock moving one step up or down, we bound the wait for useful backward moves and the rate of random erasures a stored counter tolerates. The results separate the power of feedback at equilibrium, observation from a fixed start, and the cost of obtaining and keeping information.

cs.CC

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16776

Key Relay and Secure Network Coding: One-Shot Equivalence and Separation on Undirected Networks

Yuji Ando, Edmund Chen, Go Kato, Toyohiro Tsurumaru, Derek Zhang, Donald Zvada

原文摘要

Motivated by quantum key distribution networks, we compare key relay protocols with secure network coding on the same undirected graph. In this comparison, we fix the finite field, terminal set, and wiretap family. Each edge supplies one independent uniform symbol to its endpoints for key relay, or carries one symbol once for network coding. The network code may choose its source terminal, edge orientations, and causal transmission order before execution. In this setting, the standard one-time-pad simulation readily converts every secure network code to key relay. We investigate the reverse conversion to linear network coding. For two terminals and a one-symbol key, arbitrary key relay is equivalent to scalar linear acyclic network coding over the prescribed field. The proof uses connectivity and flow conservation, related to the binary $st$-flow protocol. However, this equivalence fails in general when several terminals seek one common key. A doubled triangle separates key relay from acyclic coding but admits causal coding on a cyclic orientation. On the four-terminal complete graph, a linear key relay protocol over the ternary field withstands the three perfect-matching wiretap sets, whereas no linear causal network code distributes a source symbol of the same size, even with private randomness at every vertex and its use in decoding. We also establish a linear key relay normal form that removes interaction and private randomness and bounds public communication. All results concern one-shot feasibility with exact agreement and perfect secrecy.

cs.IT · quant-ph

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2607.13920

DeepStress: Stress-Testing Deep Search Agents

Ismael Rousseau, Geraldine Damnati, Frederic Bechet

原文摘要

While search agents demonstrate impressive capabilities in multi-step question answering, their robustness to poor-quality evidence remains under-explored. Such evidence is rare in benchmark corpora, and resulting failures are hard to reproduce. We propose DeepStress, a dynamic stress-testing framework that subjects agents to degraded evidence at controlled rates. It replaces the retrieval module with a synthetic environment where documents are generated dynamically for each query, under independent control of their trustworthiness, relevance, and factuality. To characterize behavior under stress, we go beyond binary accuracy with a five-way classification of final-answer outcomes and two new reliability-aware metrics, RAS and CoPRAS, which credit cautious behavior and account for computational cost. Evaluating 12 systems on questions from HotpotQA and BrowseComp-Plus, we find that several open-weight agents answer confidently when sources contain obvious credibility red flags, and that many HotpotQA questions are answered from parametric knowledge even when all returned documents are factually wrong.

cs.CL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16772

Class-Competence Aggregation: Output-Layer Routing for Federated Multiclass DDoS Detection under Class-Concentrated Non-IID Data

Tapu Datta, Hafiz Imtiaz

原文摘要

Federated learning over distributed intrusion-detection sensors naturally produces \emph{class-concentrated} non-IID data, where an entire attack class is held by a single client and absent elsewhere. We show this collapses federated averaging in a specific, localized way: the output layer holds one decision boundary per class, and when a class's aggregation weight comes overwhelmingly from clients that have never seen it, its neuron receives an active anti-signal rather than mere dilution, collapsing to zero. On CIC-DDoS2019, standard FedAvg drives the concentrated LDAP class to $0.00 \pm 0.00\%$ F1 across every seed. We propose Class-Competence Aggregation (CCA), which aggregates the output layer per class using each client's own locally reported validation F1 as the weight, while hidden layers remain under standard size-weighted FedAvg. CCA requires no server-held auxiliary data and adds only one F1 scalar per class per round beyond the feature-scaling and class-count exchanges common to any federated pipeline. On our primary CNN-LSTM architecture, CCA alone raises LDAP F1 from $0\%$ to $84.83 \pm 1.65\%$, and the full method reaches $91.64 \pm 0.57\%$ macro F1, closing the federated-to-centralized gap from $-42.53$ to $-4.05$ points. Benchmarked against four other federated methods under identical partitions and seeds, CCA is the only one that recovers the concentrated class substantially and reliably; a $178\times$ smaller MLP shows the same qualitative recovery, though less completely, confirming the effect holds across architecture scale even as its magnitude varies. Results are validated across three seeds and replicated on a second dataset, CICIDS2017.

cs.LG · stat.ML

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16767

Evaluating Video LLMs' Understanding of Corner Cases in Autonomous Driving

Gabriel Manalu, Minh Son Hoang, Hyeontaek Hwang, Jaesung Lee, Hyojin Choi, Tran Tuan Phong, Seon Kwon Kim, Saad Wazir 等 9 位作者

原文摘要

Natural language explanations promised by modern Video Large Language Models (Video LLMs) offer autonomous driving systems a path toward detailed scene description and interpretable decision making on the road. While recent benchmarks have begun assessing these models on driving scenes, they mainly target routine traffic conditions and general comprehension tasks based on existing driving datasets and therefore fail to evaluate model performance on the rarest and most safety-critical driving events, also known as "corner cases". To provide a comprehensive evaluation of Video LLMs on precisely these scenarios, we introduce DashBench, a benchmark testing 25 Video LLMs on 1,709 carefully curated driving videos spanning not only collision-related corner cases but also rare non-collision ones. Through extensive prompts and open-ended question answering, we expose a critical capability gap: all evaluated models perform substantially below the threshold required for safety-critical driving applications, frequently misjudging or missing dangerous events on the road ahead and hallucinating non-existent agents or actions in their explanations. We believe DashBench, combining an extensive dataset with a rigorous evaluation toolkit and a detailed diagnosis of current model limitations, drives progress toward safer, more explainable autonomous driving systems. The code and data will be shared at: cornercase-dashbench.github.io

cs.CV · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16764

Efficient Periocular Recognition through Model Scaling and Cross Spectral VIS NIR Verification

Fabian Stockhard, Juan E. Tapia, Christoph Busch

原文摘要

Periocular recognition offers an efficient alternative to full-face verification when facial regions are occluded or image resolution is limited. We present AdaEye, a systematic study of input and backbone scaling for periocular verification in visible-spectrum (VIS) and cross-spectral (VIS/NIR) settings. Using an AdaFace-style framework, we compare full-face, masked-face, and dedicated 80x40 periocular inputs while varying backbone width and embedding dimensionality. Across four VIS/VIS and VIS/NIR protocols, we assess verification accuracy, computational cost, inference throughput, training memory, template size, and training stability. Direct periocular processing reduces recognition cost from 6.30 to 1.64 GMACs per image and increases measured throughput by approximately threefold. Its effectiveness depends on the evaluation protocol: periocular models perform better on native UFPR periocular images, whereas face-based protocols favor retaining more input information. Moderate backbone width reductions maintain comparable mean verification performance on CASIA and ND-NIVL while reducing computation. Embedding compression decreases template-storage requirements, primarily benefiting deployment. Multi-seed experiments reveal that some configurations require numerical stabilization, particularly when embedding and backbone reductions are combined. These findings support selecting compact periocular architectures according to the target protocol and deployment constraints, balancing accuracy, efficiency, storage, and robustness.

cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v4arXiv:2609.10447

Compact totally separated types

Martín Hötzel Escardó

原文摘要

Perhaps surprisingly, there are infinite types that can be exhaustively searched mechanically in finite time. We use ideas from topology to build plenty of them, referring to searchable types as compact types, and we use ordinals to measure their logical complexity. We consider two systems of ordinal notations under which a single notation denotes both a discrete ordinal and a compact one, with an embedding of the former into the latter whose image has empty complement. A boolean valued function decides which points in the image of the embedding are isolated and which are topological limit points. The first system consists of the traditional Brouwer codes and the second is an inductive-recursive universe generalizing them. The discrete ordinals so obtained are trichotomous, and the compact ones have the least element property for complemented subsets, but these two desirable properties cannot be fulfilled simultaneously in a constructive setting. The compact ordinals obtained from Brouwer codes further enjoy a boolean Leibniz principle, which has the notion of total separatedness as its topological counterpart. This extends previous work from Gödel's system T to intensional Martin-Löf type theory with univalent universes, and is formalized in Agda in the TypeTopology repository.

math.LO · cs.LO · math.GN

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2509.19957

Gaze-controlled segmentation improves object recognition accuracy in simulated phosphene vision

Eleftherios Papadopoulos, Umut Güçlü, Yağmur Güçlütürk

原文摘要

Visual neuroprosthetics provide a simplified visual sense through cameras, computers, and implanted electrodes. A central challenge is deciding what to convey with the few phosphenes available. We asked whether letting the user's gaze select the object that is conveyed in full, rather than extracting edges across the whole scene, improves object recognition in simulated phosphene vision. We developed Gaze-Controlled Segmentation (GCS), in which the point of gaze prompts the Segment Anything Model and the selected object is rendered as a filled region over the scene's edges at reduced intensity. In simulated phosphene vision, eight participants performed an object search-and-localization task under GCS, edge detection, and normal-vision conditions. Both phosphene conditions were gaze-contingent, and data were analysed with Bayesian mixed-effects models. Relative to edge detection, GCS produced a modest but credible improvement in identification accuracy on target-present trials (+9.4 percentage points, 95% highest-density interval [+4.3, +14.2]). Gaze reached the target about as often under both methods, consistent with an advantage at the identification stage. GCS trials tended to be slower (+2.7 s, [-0.8, +6.6]), a difference that was not credible, and in exploratory analyses the advantage was largest for cylindrical objects and absent for round ones. Stated preferences leaned toward GCS on most usability criteria, though with eight participants these are individually consistent with chance. The comparison does not isolate the contribution of the segmentation itself. These findings illustrate the potential of gaze-driven, interactive scene representation for neuroprosthetic vision, while underscoring that the gain is modest, task-dependent, and established so far only in simulated phosphene vision with sighted observers, using phosphenes larger than current implants would produce.

cs.HC

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.00166

Early Stopping in Large Reasoning Models: A Survey and Empirical Audit of Efficiency and Reliability

Quang Hung Nguyen, Hanh Nguyen, Thanh Tam Nguyen, Quoc Viet Hung Nguyen, Trinh Pham, Thanh Trung Huynh

原文摘要

Large reasoning models can improve accuracy by generating longer reasoning traces, but some of this computation may be unnecessary. Early-stopping methods aim to stop once enough reasoning is available to answer. Evaluating these methods requires accounting for the cost of deciding when to stop and the reliability of the resulting answer. We evaluate different methods across five reasoning models on mathematical and scientific reasoning, code generation, and factuality tasks by using a shared rollout replay protocol. We introduce Target-Integrated Token Efficiency (TITE) to measure efficiency relative to the oracle across accuracy targets. We also examine correctness at later exit positions, transfer across models and tasks, post-stop reasoning continuation, and retained justification quality. In our experiments, correctness can decrease at later positions beyond what the fitted decoding-noise model predicts. Reliability checks also reveal continued reasoning after forced termination and weaker supporting reasoning in some correct early-stopping outputs. The implementation can be found in https://github.com/hung20gg/audit-early-stopping.

cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16759

Propositional Dynamic Logic without tests does not have Craig Interpolation

Malvin Gattinger

原文摘要

It was shown recently by Borzechowski et al. (2026) that Propositional Dynamic Logic (PDL) has Craig Interpolation. Whether test-free PDL, i.e. PDL without the `?' operator, has interpolation remained open. Here we answer this question in the negative with a counterexample: a test-free PDL validity for which no test-free interpolant exists. The validity only uses one atomic program and the Kleene star, so it implies that the basic modal logic K extended with the reflexive transitive closure does not have interpolation, as previously shown by Studer (2009). The example was originally obtained using LLMs and Lean. We describe our methodology and provide links to all prompts, LLM output summaries as well as the Lean proof.

cs.LO · math.LO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16758

Tracking relevance syntactically: A nested approach to weak relevant logic

Fabio De Martin Polo, Han Gao

原文摘要

The variable-sharing property (VSP) is regarded as a key feature of relevant logics, while the Craig interpolation property (CIP), formulated with respect to relevant implication, implies the VSP in constant-free propositional languages. However, to the best of our knowledge, no constant-free distributive relevant logic has previously been shown to enjoy the CIP. This absence of positive results has contributed to a widespread impression among relevant logicians that relevance and Craig interpolation are incompatible. In this work, we address this tension by proving that the negation-free fragment of base relevant logic B, denoted B+, enjoys both properties. Our proof is purely syntactic and is based on a novel nested sequent calculus featuring binary relevant blocks as primitive structural constructs. Unlike conventional nested sequent calculi, where unary nesting reflects the binary accessibility relation of Kripke semantics, our calculus employs a binary nesting operator that directly reflects the ternary relational structure of Routley-Meyer semantics for relevant logics. After establishing soundness and cut-free completeness of the calculus, we prove several meta-logical properties for the logic. Most notably, we provide the first syntactic proof of the VSP (along with a variety of its stronger variants) and the first proof of the CIP for B+, both derived directly from the cut-free system using a structural notion internalized within nested sequents.

cs.LO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16756

Beyond Binary Classification: Multiclass Machine Learning for Diabetes and Prediabetes Risk Prediction Using NHANES

Aman Mujahid, Anita Penkova

原文摘要

Detecting diabetes risk early is important for preventing disease progression. Most existing machine learning approaches focus on binary classification, which limits their ability to capture intermediate risk states such as prediabetes. We develop a supervised learning framework to classify adults as normal, prediabetic, or diabetic using National Health and Nutrition Examination Survey data. The framework uses routinely collected demographic, clinical, and lifestyle variables from 15,356 adults, with glycated hemoglobin (HbA1c) used only to define labels and excluded from input features to prevent leakage. We compare conventional machine learning and transformer-based models and examine two binary formulations: normal versus prediabetes, and normal versus combined prediabetes and diabetes. Models are evaluated on a held-out test set using precision, sensitivity, specificity, and F1 score. In the multiclass setting, performance is broadly consistent across models, with accuracy up to 0.72 and macro-averaged F1 of 0.70. Prediabetes remains the most challenging class, with F1 scores of 0.54 to 0.56 for conventional models and 0.53 to 0.55 for transformers, reflecting its overlap with both neighboring states. When the task is reduced to normal versus prediabetes, F1 scores rise modestly to 0.62 to 0.64. A more substantial improvement is observed under a broader screening formulation combining prediabetes and diabetes, with F1 scores of 0.73 to 0.74 across all models. These findings suggest that feature overlap along the biological continuum between normal and diabetic states contributes more to the difficulty of prediabetes detection than model architecture. Multiclass prediction provides a more clinically informative representation of metabolic risk progression but highlights a trade-off between clinically meaningful stratification and predictive accuracy.

cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16755

A Polynomial-Time Sixth-Power Excluded Grid Theorem

Mohammad Hajiaghayi

原文摘要

We give a polynomial-time construction for a sixth-power excluded-grid theorem. Every graph whose treewidth is at least $g^6$ times a fixed polylogarithmic factor contains a $g\times g$ grid minor that our randomized algorithm finds with constant probability. The algorithm outputs an explicit minor model, and the exponent in its running time is independent of $g$. This improves the ninth-power existence bound of Chuzhoy and Tan~\cite{CT21} and the $g^{98}$ threshold in the polynomial-time construction of Chekuri and Chuzhoy~\cite{CC16}. Quantitative bounds relating treewidth to grid size are central to bidimensionality, developed by Demaine, Hajiaghayi, and collaborators~\cite{DH05,DH08}. The main ingredient is an asymmetric local routing theorem that uses substantially fewer attachment paths than backbone paths. A balanced recursion preserves connectivity by charging deletions to the boundary of a fixed reference cluster. We construct the smaller attachment sets through a routing hierarchy: either it exposes an accessible well-linked boundary, or its quotient graphs contain the desired grid. These arguments implement cluster splitting using approximate cuts and flow computations. An explicit cut-matching construction and a constructive grid-extraction procedure complete the algorithm. We also give a general packing--covering hypothesis for trees meeting three prescribed terminal sets under which the exponent can be reduced below six. A companion Lean development verifies the new construction, its success guarantee, and the composition of its polynomial running time, using explicit interfaces for the published algorithms and standard finite-data operations.

cs.DS

查看论文
arXiv ↗PDF ↗
版本更新 · v4arXiv:2609.39329

PatchKV: Weight-Space Compensation of KV Cache

Chanryeol Lee, Chanhyuk Lee, Yeonwoo Choi, Donggyun Kim, Seunghoon Hong

原文摘要

Long-context inference with Large Language Models (LLMs) is bottlenecked by the linearly growing memory of the key-value (KV) cache. Existing compression methods reduce the cache through token eviction or approximation, but degrade sharply at aggressive compression budgets. We propose PatchKV, a training-free framework that compensates KV cache compression methods by carrying part of the context in the model's weights. PatchKV pairs an off-the-shelf compressed KV cache with a context-specific weight patch, which is computed once at context-loading time and served for downstream queries for the context. The weight patch is derived in closed form via ridge regression, by aligning the block-wise activations of context-derived reference query tokens under the full cache and the compressed cache. Once merged into the model, the patch leaves the forward graph and per-query inference cost unchanged in the single-context, multi-query setting. Across long-context QA (SCBench with up to 170K tokens, SQuAD, NIAH) and math (GSM8K) benchmarks on three model architectures, PatchKV consistently improves cache compression methods, suggesting an alternative direction to compensate them at aggressive budgets. Code is available at https://github.com/cusasak/PatchKV.

cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16752

AgriSense: Vision-Language Decision Support and Robotic Intervention in a Virtual Greenhouse Copy title

Lyes Saad Saoud, Oualid Doukhi, Reza Ghorbani, Moussa Ayyash

原文摘要

Autonomous greenhouse management increasingly requires robotic systems that integrate perception, navigation, environmental regulation, and plant-level intervention. This paper presents AgriSense, a framework combining an Isaac Sim virtual greenhouse, a quadruped mobile manipulator, RGB and LiDAR-inertial sensing, environmental control, and supervised vision-language decision support. Frozen DINOv2 and BERT encoders, together with structured environmental inputs, support joint prediction of twelve environmental references, a target zone, and an intervention flag. Checkpoint replay reproduces 41/43 zone decisions. Across five seeds on a source-grouped test set, the full, image-removed, and language-removed configurations each achieve 43/44 correct zone decisions, providing no evidence of incremental multimodal zone-accuracy improvement. The full model achieves mean absolute errors of 0.249 degrees C for temperature, 2.030 percentage points for humidity, and 0.997 percentage points for soil moisture. Sensor-only inputs match full-model zone accuracy, while language-only inputs achieve 68.64%. External real-image benchmarks reveal distribution-shift limitations. Five offline Point-LIO routes yield rigidly aligned trajectory errors of 5.22-17.32 mm. Of 40 virtual-arm targets, 32 satisfy geometric acceptance criteria and eight are rejected at boundaries. Six long-horizon climate cases and fault protocols characterize integration limits. Navigation and integrated treatment rely on simulator-state inputs; physical greenhouse performance and agronomic treatment efficacy remain unvalidated. The contribution is an evaluated decision-support and robotic-intervention framework for a virtual greenhouse.

cs.RO · cs.AI · cs.CV

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16750

Interpretable machine learning diagnosis of degradation mechanisms in lead-free $RbSn_{0.5}Ge_{0.5}I_{3}$ solar cells from temperature- and illumination-dependent electrical measurements

Nabanita Sarker, Hafiz Imtiaz

原文摘要

Machine learning coupled with device simulation is widely used to optimize thin-film solar cells, but almost exclusively in the forward direction, from device parameters to predicted performance. The inverse problem encountered after fabrication, namely identifying the mechanism that limits a given device, is still addressed by manual inspection of characterization data. In this work, both directions are examined for lead-free $RbSn_{0.5}Ge_{0.5}I_{3}$ solar cells using approximately 42,000 SCAPS-1D simulations. Replacing the Spiro-OMeTAD hole transport layer with CuI increases the simulated power conversion efficiency from 19.6\% to 21.6\%, almost entirely through the fill factor. Surrogate models trained and tested on a factorial parameter grid substantially overstate their predictive accuracy. An XGBoost regressor attains $R^2 = 0.9997$ on held-out grid points but only 0.58 on independently sampled devices, whereas training on a Latin hypercube design of comparable size restores $R^2 = 0.995$. For the inverse problem, a multi-label classifier identifies which of six degradation mechanisms are present from simulated temperature- and illumination-dependent open-circuit-voltage measurements. It achieves a macro $F_1$ score of 0.938, compared to 0.687 for a calibrated physics-rule baseline. The classifier does not, however, generalize to combinations of mechanisms absent from the training set.

cs.LG · math-ph

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2605.13385

Exact Accepting-State Spectrum for Reversal of Permutation Automata

Samuel German

原文摘要

In this paper we determine the accepting-state spectrum of the reversal operation on permutation automata, and moreover prove the conjecture conjectured by Rauch and Holzer in so doing. Namely, we show that reversal satisfies the spectrum $g^{\operatorname{asc}}_{R,\mathrm{PFA}}(0)=\{0\}$, $g^{\operatorname{asc}}_{R,\mathrm{PFA}}(1)=\{1\}$, and $g^{\operatorname{asc}}_{R,\mathrm{PFA}}(m)=\mathbb{N}_{\ge 2}$ for $m\ge 2$. We do this by, for every $m\ge 2$ and every $α\ge 2$, constructing a binary permutation automaton $A_{m,α}$ such that $ \operatorname{asc}(L(A_{m,α}))=m$ and $ \operatorname{asc}(L(A_{m,α})^{R})=α$. The spectrum above follows from this (combined with the trivial cases $m=0$ and $m=1$ and the fact that $1$ is magic for every $m\ge 2$). Thus our result shows that for $m \ge 2$, the reversal operation in permutation automaton attains ever single value positive value in the accepting-state complexity except for $1$.

cs.FL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16749

ORCA:Orthogonal Residual Consensus Alignment for Multi-View Clustering

Xiaojian Ding, Xian Li, Xiaoying Zhu, Lin Zhao, Kaixiang Wang

原文摘要

Deep multi-view clustering often suffers from redundant information accumulation during fusion and indiscriminate alignment across representation levels. We propose Orthogonal Residual Consensus Alignment (ORCA) to address these challenges. Its semantic incremental recurrent fusion module extracts cross-view orthogonal residuals as non-redundant increments and adaptively weights recurrent states through divergence-guided gating. Its consensus-calibrated hierarchical contrastive alignment module uses the global consensus to weight contrastive alignment of view-level, pairwise, and recurrent representations. Together, these modules balance view-specific structure with cross-view consistency. Experiments on five benchmarks against fifteen competing methods demonstrate ORCA's effectiveness, including a 5.25 percent accuracy improvement over the strongest baseline on 100leaves. Ablation studies validate both modules. Code is available at https://github.com/XiaojianDing/2026-NeurIPS-ORCA.

cs.CV · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.16748

PhysLoop: Harnessing LLMs' Physics Awareness for Self-Improving Video Generation Models

Yuxuan Chen, Xiaojie Li, Hao Chen, Juncheng Ma, Kewei Zhang, Zhongjing Du, Guoqiang Wei, Fangfang Wang 等 14 位作者

原文摘要

Video generation models produce visually realistic videos but still struggle with physical dynamics and causal consistency. Learning from generated videos requires fine-grained physical supervision, yet reliable quality judgments, correction pairs, and controlled variations are difficult to obtain. Large language models (LLMs) possess physical knowledge, but struggle to reliably guide video generation models in producing their training videos. We present PhysLoop, a unified harness designed to support physically grounded self-improvement in video generation models. The harness guides and constrains LLMs throughout video generation, selection, and correction, distilling their physical knowledge into physically consistent videos. The framework supports controllable training-data construction and inference-time video repair. By identifying and correcting failures, PhysLoop marks a key milestone toward recursive self-improvement (RSI) in video generation models with limited human intervention. We release 100K videos spanning diverse, realistic indoor and outdoor scenes, along with generation tools and two evaluation datasets: PhysLoop-Gen and PhysLoop-Repair. We evaluate controllable generation and localized repair on these datasets, respectively, and assess physical behavior on the external Physics-IQ Verified benchmark.

cs.CV · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.34924

Audit the Scaffold, Not the Checkpoint: A Stationarity Dichotomy for Recursive Self-Improvement in Agentic Coding

Sebastian Bobadilla-Suarez, Bob Suh, Ryan Fortin

原文摘要

Rules for AI self-improvement often key on the model itself: has it been retrained, or are its weights frozen? We argue that, for AI coding agents, this check is incomplete, and we propose, but do not run, the experiment that would test it. An agent that keeps revising its own work runs into diminishing returns as long as the set of changes it can reach stays fixed, and can escape them only if that set grows. Our hypothesis is that the set can grow with the model untouched, because the software around the model, its scaffold of tools, tests and task breakdown, can be rewritten. If so, a frozen model guarantees a ceiling eventually but not steady behavior on the way there, and audits should track changes to the scaffold, not only to the model. The same criterion separates three things often lumped together: more search within a fixed setup, retraining the model during use (which can raise the ceiling), and rewriting the scaffold, which sits between them. Adding more agents does not lift the ceiling. Picking the best of several agents' answers reaches the best single agent's ceiling exactly, and only gets there faster. A majority vote could do better, but only if the agents fail on different tasks. In our tests, 30 agents from one model family failed together: some task defeated all of them, and a majority failed 23/55 (42%) of tasks. Our evidence is descriptive. On a public coding benchmark, the gain from each extra round of revision shrinks toward zero, though the unsolved tasks left to measure are also fewer and harder. Across 401 development sessions, the amount of code changed falls off from commit to commit, but programmers working before AI tools show the same pattern, so the data show that current practice levels off without showing why. We reach the criterion by treating repeated revision as a known machine-learning method, boosting, and note where real agents depart from it.

cs.LG · cs.AI · cs.SE

查看论文
arXiv ↗PDF ↗