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

+406新增论文
+306其中 AI 相关
+55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:2026-11-16 11:42各信源进度 →

统计口径与数据新鲜度

统计 2026-11-15 12:05 — 2026-11-16 12:05(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:2026-11-16 11:42库内论文最新更新时间:2026-11-14 01:29

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。

23,754 已收录论文19,886 AI 相关6,892 有修订112 HF 社区发现

浏览论文库

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 23,754 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:2611.14823

Discharge Summary Generation Is Not Just Summarization: A Reality Check for Evaluation

Radhika Dua, Frederick Wolff, Kyunghyun Cho, Eric Karl Oermann

原文摘要

Discharge summaries combine a retrospective account of hospitalization with discharge-time decisions. Physician-written references therefore contain both facts supported by prior clinical notes and reference-only content, potentially reflecting clinical synthesis, handoffs, or discharge planning. Reference-only evaluation can conflate failure to recover available information with failure to reproduce content absent from the model input. We introduce FairSumEval, an evidence-aware, fact-level framework that assesses whether each reference and generated fact is supported by the model's evidence and captured by the other summary. The resulting outcome map separates matches, omissions, and additions. Two complementary metrics, Soft IoU-S and Soft IoU-NS, measure reference-generation agreement for facts with and without identified evidence support. On 200 MIMIC-III admissions, 18.8% of physician reference facts have identified support in the prior evidence, compared with 65.4% and 63.5% of facts generated by a general-purpose and a medically specialized model. Overlap is substantially higher for supported content (0.255 and 0.236) than for content without identified support (0.030 and 0.026). The two scores are weakly correlated across admissions (rho = 0.22 and 0.17), and conventional metrics do not consistently capture both dimensions. FairSumEval distinguishes evidence-grounded summarization from agreement on content without identified support, revealing what models omit and add in each category.

cs.CL · cs.AI · cs.LG

版本更新 · v2arXiv:2605.31547

The Dynamic-Probabilistic Consistency Gap in Chaotic Surrogate Modeling

Andre Herz, Matthijs Pals, Daniel Durstewitz, Georgia Koppe

原文摘要

Dynamical systems reconstruction (DSR) aims to learn surrogate models that capture the dynamics underlying time-series data. Reliable scientific inference with these surrogates requires uncertainty estimates consistent with the learned dynamics. We expose a dynamic-probabilistic consistency (DPC) gap: optimizing finite-horizon probabilistic objectives can degrade the learned dynamics or decouple predictive uncertainty from the local tangent dynamics it should reflect. We isolate three mechanisms behind this gap: core collapse, noise masking, and blind uncertainty. Specifically, we show that open-loop rollout objectives can penalize covariance growth driven by the dynamics, encouraging optimization shortcuts that weaken local expansion. To mitigate this gap, we propose KAFFEE (Kalman-Aware Framework For Ergodic Emulation), a differentiable extended Kalman filter-based training framework that jointly learns dynamics and noise covariances by coupling filtered innovation likelihoods with covariance propagation through learned local Jacobians, while differentiating through the Kalman updates. On stochastic hyperchaotic Lorenz-96, KAFFEE mitigates the identified failure modes and improves both reconstruction of dynamical invariants and predictive scores relative to open-loop objectives. The DPC gap also appears when probabilistically adapting a DSR foundation model across 13 chaotic systems, where KAFFEE improves predictive likelihood over open-loop adaptation while preserving the pretrained model's long-term dynamics and enabling in-context Bayesian filtering.

cs.LG · math.DS · stat.ML

新论文 · v1arXiv:2611.14822

Quantum Telatar formula: fading quantum capacity of Haar-random multimode attenuators

Eric Culf, Victor V. Albert

原文摘要

Cellular communication is done over an imperfectly predictable multimode environment, where the sender only knows about the distribution of possible intermode coupling and scattering a signal faces. It may thus seem surprising that such communication is efficient. Telatar proved that the sender's lack of exact knowledge about the channel, on average, does not significantly hinder multimode communication. As the number $N$ of modes increases at fixed signal-to-noise ratio, the average capacity of a classical multimode fading channel grows linearly in $N$, like that of $N$ decoupled single-mode channels. We study the quantum analogue in a Haar-random passive bosonic network and prove a similar scaling for unassisted quantum information transmission. Suppose that the sender drives a fixed fraction of the input modes and the receiver collects more than half of the output modes. Then the quantum capacity of such channels at zero temperature grows linearly with the number of modes at fixed photon number per mode, just like the classical capacity grows with the number of antennas. At nonzero temperature, the Haar-averaged coherent information of the isotropic thermal input bounds the capacity from below, and its value per mode converges to an integral against Wachter's law. We numerically identify positive achievable rates in models motivated by cryogenic microwave networks and room-temperature multimode fibers.

quant-ph · cs.IT

版本更新 · v2arXiv:2605.11872

LOFT: Low-Rank Orthogonal Fine-Tuning via Task-Aware Support Selection

Lanxin Zhao, Bamdev Mishra, Pratik Jawanpuria, Lequan Lin, Dai Shi, Junbin Gao, Andi Han

原文摘要

Orthogonal parameter-efficient fine-tuning (PEFT) adapts pretrained weights through structure-preserving multiplicative transformations, but existing methods often conflate two distinct design choices: the subspace in which adaptation occurs and the transformation applied within that subspace. This paper introduces LOFT, a low-rank orthogonal fine-tuning framework that explicitly separates these two components. By viewing orthogonal adaptation as a multiplicative subspace rotation, LOFT provides a unified formulation that recovers the multiplicative core of representative orthogonal PEFT methods, including coordinate-, butterfly-, Householder- (with a learnable support) and principal-subspace-based variants (up to scaling and relaxation). More importantly, this perspective exposes support selection as a central design axis rather than a byproduct of a particular parameterization. We develop a first-order analysis showing that useful adaptation supports should be informed by the downstream training signal, motivating practical task-aware support selection strategies. Across language understanding, visual transfer, mathematical reasoning, and multilingual out-of-distribution adaptation, LOFT with task-aware supports improves performance at matched trainable-parameter and transform settings under a shared protocol. This suggests that, under a fixed transform and budget, the choice of support affects performance, making systematic support selection an important direction for improving orthogonal PEFT.

cs.LG · stat.ML

新论文 · v1arXiv:2611.14821

From Classification to Generation: Vision-Language Models for Roman Imperial Coin Description

Pedro Lourenco, Ognjen Arandjelović

原文摘要

Cataloguing a Roman Imperial coin involves reading its legends and identifying its imagery in order to match it to an entry in the standard reference catalogue. Previous work has automated individual parts of this process, such as identifying the issuer or recognizing reverse motifs, by formulating each part as a classification task over a predefined set of labels. We instead formulate coin description as image-to-text generation and adapt an open-weight vision-language model (VLM), Qwen3-VL-8B-Instruct, to four tasks: transcribing the obverse and reverse legends, and describing the obverse and reverse imagery. To obtain training data, we convert approximately 30,000 auction records into structured image-text pairs. We separate the two sides of each coin, extract and normalize legends and descriptions, and standardize obverse descriptions against a bust-type taxonomy. We then train a separate low-rank adapter for each task on a single 20GB GPU, and evaluate each task with metrics that we define to capture the key components of numismatic descriptions. The adapted models outperform both the unadapted base model and GPT-4.1 on all four tasks, reducing the normalized edit distance of legend transcriptions by over 40% relative to GPT-4.1. The improvement is smaller, but still present, on legends absent from the training data.

cs.CV

新论文 · v1arXiv:2611.14818

What Do Tokenizer Comparisons Measure? Disentangling Granularity, Root Assignment, and Word-Level Readout across Amharic and Hausa

Abdul Rehman Antall, Baidya Nath Saha

原文摘要

Morphology-aware tokenizers are expected to help morphologically rich languages, and their reported gains over byte-pair encoding (BPE) are usually credited to their linguistic units. A comparison with BPE cannot support this attribution, since such a tokenizer also segments text more finely (granularity) and, on token-level tasks, changes how much the sub-token readout matters. We separate both factors from the effect of grouping words by their roots (root assignment) in Amharic, a Semitic language whose Ge'ez abugida hides consonantal roots inside syllabic glyphs, with Latin-script Hausa as an orthographic contrast. Our design combines a lossless root-and-pattern tokenizer with controls that change one factor at a time, in two independent implementations. In both, real root assignment improves Amharic named-entity recognition over scrambled roots (+4.9 and +4.7 F1), and root-frequency structure explains little of this gain. In small from-scratch models, the readout choice moves F1 by up to 13.4 points and reverses the sign of every comparison of the root-and-pattern arms with BPE, whereas in two pretrained encoders it moves F1 by less than one point. The granularity effect changes sign across languages and implementations. We recommend decomposing tokenizer effects, and reporting the readout, before a gain is attributed to linguistic structure.

cs.CL · cs.LG

新论文 · v1arXiv:2611.14817

Tripod Reordering: A 33-Queue Layout of Planar Graphs

Sergey Pupyrev

原文摘要

A queue layout of a graph consists of a linear order of the vertices and a partition of the edges into queues so that no two edges in a single queue are nested. The minimum number of queues needed in a queue layout of a graph is called its queue number. A planar product structure theorem of Dujmović, Joret, Micek, Morin, Ueckerdt, and Wood (JACM, 2020) states that every planar graph is a subgraph of the strong product of a graph of simple treewidth at most $3$, a clique $K_3$, and a path. The result has been used to show that the queue number of planar graphs is at most $49$. Recently, Bekos, Gronemann, and Raftopoulou (Algorithmica, 2023) investigated how the general approach based on the product structure can be optimized for planar graphs. They claim that an appropriate reordering of the three vertices associated with each bag in the product structure decomposition, called a tripod, reduces the queue number of planar graphs to $42$. We first highlight a gap in their queue layout algorithm: one of the choices required by the algorithm is not guaranteed to exist. Then we show how to address the limitation and provide an alternative tripod reordering mechanism that reduces the upper bound on the queue number of planar graphs to $33$. As a corollary, bipartite planar graphs have queue number at most $22$.

cs.DM · cs.DS · math.CO

版本更新 · v2arXiv:2605.17757

OSCAR: Offline Spectral Covariance-Aware Rotation for 2-bit KV Cache Quantization

Zhongzhu Zhou, Donglin Zhuang, Jisen Li, Ziyan Chen, Shuaiwen Leon Song, Ben Athiwaratkun, Xiaoxia Wu

原文摘要

INT2 KV-cache quantization is attractive for long-context LLM serving, yet making it both accurate and practical remains challenging. Data-oblivious rotations such as Hadamard transforms flatten channel outliers, but still degrade sharply at INT2 because they ignore how downstream attention queries the cache. We propose OSCAR, an ultra-low-bit KV-cache quantization method that estimates attention-aware covariance targets offline and uses them to derive fixed per-layer rotations and clipping thresholds. By aligning the quantization basis with the second-moment structures consumed by attention, OSCAR preserves attention fidelity at 2-bit precision. Beyond theoretical analysis under a frozen-error surrogate, we build a serving-ready OSCAR system with fused Triton write and decode kernels that preserve paged KV-cache layouts and prefix caching, integrating into modern serving engines such as SGLang and vLLM. We evaluate OSCAR on reasoning and coding benchmarks with generation traces of up to 32K tokens across five tasks. On Qwen3-4B-Thinking-2507 and Qwen3-8B, OSCAR narrows the BF16 accuracy gap to 3.78 and 1.21 points, respectively, where Hadamard-only INT2 collapses, and matches BF16 within run-to-run variance on Qwen3-32B and GLM-4.7 (358B parameters). On RULER-NIAH up to 128K tokens, OSCAR substantially outperforms QuaRot-style INT2 across context lengths. System-wise, OSCAR compresses KV-cache memory by nearly $8\times$, boosts multi-request throughput by up to $7.8\times$ at large batch sizes within the same memory budget, and accelerates batch-size-1 decoding by up to $3.1\times$ over BF16 by cutting KV memory traffic.

cs.LG · cs.AI · cs.DC · cs.PF

新论文 · v1arXiv:2611.14815

BirdDreams: Joint Perspective and BEV Video Generation for a Plausible, Reactive Driving World Model

Junyoung Seo, Jaewoo Seo, Seungryong Kim, Laura Leal-Taixé, Xiuming Zhang, Seung Wook Kim

原文摘要

Driving world models aim to simulate how a scene unfolds under different ego actions. This requires plausible motion of surrounding vehicles as well as precise ego-trajectory control. Learning vehicle motion from perspective video alone is challenging because motion is entangled with appearance and vehicles can be occluded. Prior work therefore conditions video generation on layouts that specify vehicle motion, providing control over surrounding vehicles but requiring their future trajectories as inputs. We introduce BirdDreams, an autoregressive driving world model that jointly generates perspective video and a schematic bird's-eye-view (BEV) video, making scene layout part of the predicted future. The BEV, rendered from scene annotations for training, depicts roads and vehicles as simple shapes at a fixed metric scale, providing a top-down view of their positions and motion with reduced occlusion between vehicles. Perspective video, by contrast, retains scene details that the BEV omits, such as traffic signals and object appearance. Joint attention between two video diffusion transformers, one per stream, lets each stream inform the other during generation. The predicted BEV also provides a geometric interface for ego control: we translate and rotate its latest frame along an input ego trajectory and use the resulting frames as a soft condition that shows how the scene would appear to move if only the ego vehicle moved. At inference, BirdDreams generates both videos from a single camera image without requiring an HD map or vehicle-state inputs. On PhysicalAI-AV and nuScenes, BirdDreams follows input ego trajectories more faithfully than prior driving world models and generates more plausible ego motion without ego-action conditioning. Human evaluation shows that BirdDreams outperforms a perspective-only baseline in generating realistic surrounding-vehicle behavior.

cs.CV

版本更新 · v3arXiv:2610.29873

HAWK: Heterogeneity-Aware KV Cache Transfer for Efficient GPU-NPU Phase-Disaggregated LLM Inference

Junhyeok Park, Jungeun Yoon, Kyungmin Kim, Hunsoon Lee, Shinyoung Ahn, Kangho Kim

原文摘要

Large language model (LLM) inference exhibits distinct computational characteristics across its two phases: prefill is compute-intensive, whereas decode is more memory-bound. As a result, AI accelerators achieve varying levels of efficiency across the two phases. GPUs are well suited to prefill, but are often underutilized during decode. By contrast, NPUs can utilize their compute resources more effectively during decode. Motivated by this observation, we explore a heterogeneous architecture that integrates GPUs and NPUs, mapping prefill to GPUs and decode to NPUs to achieve high performance and efficiency. However, realizing such heterogeneous integration is challenging because LLM phase disaggregation requires KV cache transfer from prefill to decode (that is, from GPUs to NPUs), which is complicated by differences in inter-device communication mechanisms and KV cache tensor layouts. To address this, we propose HAWK, a transfer mechanism that enables rapid KV cache transfer across heterogeneous devices. HAWK establishes a remote direct memory access (RDMA) path for direct cross-device transfers while handling inter-device tensor layout discrepancies as part of the transfer process. In particular, HAWK allocates tensor-layout-aware RDMA buffers on the NPU side to match the target tensor layout. Leveraging these buffers, HAWK orchestrates data transfer from a conventional contiguous GPU RDMA buffer to the corresponding NPU buffers, allowing transferred tensors to be consumed directly without additional data rearrangement while maintaining transfer efficiency. We implement and evaluate HAWK on a system built with commercially available GPUs and NPUs, demonstrating performance within 7.88% of a GPU-only system under the same SLO constraints while achieving up to 1.26$\times$ higher power efficiency and 1.72$\times$ higher cost efficiency.

cs.AR · cs.DC

新论文 · v1arXiv:2611.14811

Improved last-iterate bounds for SGD with and without replacement

Yimin Zhong, Jiaxin Zhou

原文摘要

In this work, we study last-iterate convergence bounds for stochastic gradient descent (SGD) on convex and $β$-smooth loss functions and improve the bound from $\cO((\log T)^3/T)$ to $\cO((\log T)^3/(T\log\log T))$ in the interpolation regime. The improvement comes from applying a horizon-dependent multiscale estimate of gradients to the comparison of trajectories. The bound applies to both sampling with replacement and sampling without replacement within one pass.

stat.ML · cs.LG · math.OC

新论文 · v1arXiv:2611.14808

OverServe: Resource Amplification Attacks on SaaS Agents via Tool--Prompt Co-Design

Ruiheng Yu, Oteo Mamo, Siyu Gan, Weikuan Yu, Mulong Luo

原文摘要

Advances in large language models (LLMs) are enabling software-as-a-service (SaaS) providers to automate longer business workflows with agents. In subscription-based SaaS deployments, model and tool calls consume provider resources without additional customer payments. Because the agent autonomously decides when to stop, a malicious tenant can send a prompt to the agent, exploiting its completion decisions to extend execution and waste provider resources. In this paper, we present OverServe, a multi-agent framework for constructing such attacks against SaaS agents. OverServe co-designs Model Context Protocol (MCP) status tools and completion-confirmation prompts for the attacker. Each tool call returns accurate status and terminates normally, yet the agent may repeatedly check an unfinished condition to satisfy the prompt. This consumes the SaaS provider's resources even after the original business task is completed. OverServe constructs service-specific tool--prompt pairs and uses separate tool and prompt reviews to guide candidate pair selection. Tested tool specifications trigger no alerts in the evaluated MCP scanners. We evaluate Qwen, Nemotron, DeepSeek, and Sonnet on selected tasks from nine benchmarks. All OverServe runs pass the original task checks. OverServe can reach a maximal amplification of API requests by $93.88\times$ while preserving the original task goal and collateral checks.

cs.CR · cs.AI · cs.SE

版本更新 · v2arXiv:2611.11714

Steering Emergent Behavior in Multi-Agent LLM Systems via Information Networks

Sepehr Ilami, Babak Heydari

原文摘要

Large language model (LLM) agents increasingly operate in groups, each pursuing an individual objective while jointly determining system performance. If an operator controls which agent observes which other agent, but not their payoffs or messages, how should observations be structured to raise collective welfare? We make the information network (who observes whom) a design object, decoupled from the fixed interaction network over which agents act. Ten identical LLM agents play repeated group-decision games; by default each observes only an anonymous summary of the group's previous-round outcome, and a manager with a fixed budget adds directed observation links that reveal peers' past actions. Across 1,080 episodes and six games spanning coordination, anti-coordination, social dilemmas, and discovery, we ask how much information to share and, at a fixed budget, who should observe whom. Disclosure creates behavioral coupling: observation acts as a contagion channel whose welfare effect depends on the alignment between individually attractive behavior and the collective objective. First, the value of information is alignment-contingent: more peer information tends to help or be neutral when individual behavior is aligned with the collective goal, but under free-riding or congestion externalities it amplifies collectively harmful behavior; in a public-goods game the group does best with no information at all. We read this sign separation as a cross-game regularity within one controlled LLM population, not a model-invariant law. Second, at a fixed budget, who observes whom matters primarily when coupling is harmful: a modular "barbell" that lets sub-populations entrench harmful behavior is reliably the worst wiring, reversing the network-reciprocity intuition from human evolutionary game theory. These results argue against a default of full transparency for agent collectives.

cs.MA · cs.AI · cs.GT · cs.SI

新论文 · v1arXiv:2611.14804

GenPFN: An In-Context Tabular Foundation Model Approach for Augmenting Small Datasets in Engineering

Noah J. Bagazinski, Cyril Picard, Faez Ahmed

原文摘要

Costly simulations and experiments limit the data available for engineering design. This paper proposes GenPFN, a training-free candidate generator guided by a pretrained tabular Prior-Data Fitted Network (TabPFN) classifier with fixed weights. The context contains evaluated designs labeled desired (Class 1) or undesired (Class 0). GenPFN interpolates between desired designs and refines candidates using gated classifier gradients and Gaussian noise. Each candidate is evaluated using the problem's constraints and objective functions. Evaluated desired designs and their objective values augment the context for regression training. Evaluation spans fourteen selected problems, including a two-parameter benchmark and a 44-parameter hull problem. In paired comparisons, GenPFN with TabPFN v1 achieves higher class satisfaction and lower realism distance than the tested adversarial, variational-autoencoder and diffusion generators. The variational-autoencoder generator provides greater coverage and diversity. Relative to the synthetic minority over-sampling technique (SMOTE), GenPFN provides lower coverage distance and greater diversity. Among observed paired regression fits, augmentation reduces estimated AutoGluon validation mean squared error by 62.1% relative to context-only training. A hull-stern case study yields 99.3% propeller-compatible candidates versus 1.4% among randomly sampled context designs. Generation followed by evaluation produces sample sets that augment limited datasets and support training more accurate predictive models.

cs.AI

版本更新 · v2arXiv:2606.18532

AI Sandboxes: A Threat Model, Taxonomy, and Measurement Framework

Inderjeet Singh, Haitham Mahmoud, Andrés Murillo

原文摘要

AI sandboxes produce evidence about systems that may act through software tools, networked devices, or physical actuators. The interpretation of that evidence depends on the boundary of the experiment. We define a sandbox apparatus, distinguish five overlapping archetypes, and specify fifteen claim-relative measurement dimensions. A weakest-link licensing rule caps a deployment claim at its weakest necessary evidence dimension. We connect this rule to a threat model covering both outward harm and corruption of the evaluation apparatus, then apply it to four worked examples spanning driving, industrial control, tool-using agents, and vision-language-action manipulation. We also map the required artifacts to standards and regulatory instruments. The framework identifies which claims a bounded experiment supports, which failures its controls contain, and which assumptions require further validation before deployment.

cs.CR · cs.AI · cs.RO · cs.SE

版本更新 · v2arXiv:2610.13292

Recovering Irreducible Three-Body Interactions with Lagrangian Neural Networks

Sena Kalabalik, Sukru Caglar, Abdullah Umut Hamzaogullari, Arkadas Ozakin

原文摘要

Lagrangian Neural Networks (LNNs) learn mechanical systems from trajectory data by parameterizing a scalar Lagrangian and deriving the dynamics through the Euler-Lagrange equations, biasing the model toward motion that respects the variational structure of mechanics. The interactions studied so far have been pairwise-additive, leaving open whether LNNs can recover an irreducible many-body interaction from motion alone. We address this using the Axilrod-Teller triple-dipole interaction, the leading three-body dispersion term, whose dependence on the collective geometry of three particles cannot be represented by any sum of pairwise terms. We prescribe the kinetic energy exactly and learn only the potential, as a function of interparticle distances and angles. Combined with a signal-aware sampling strategy that concentrates training data where the three-body signal is strongest while preserving broad coverage of configuration space, this recovers the interaction from accelerations alone. The learned potential reproduces the radial scaling, angular dependence, and characteristic sign reversal of the Axilrod-Teller interaction, and generates stable trajectories. Embedding it within a pairwise Lennard-Jones potential lets us examine its observability: the three-body term is recovered once its dynamical signature exceeds the residual modeling error, which it does cleanly at amplified coupling but not at the coupling of a real noble gas, bracketing the transition between the two. Finally, symbolic regression on the trained potential, with no assumed functional form, returns the Axilrod-Teller expression with angular numerator, radial law, and coupling constant intact. An LNN with domain-specific structure can thus recover an irreducible three-body interaction from trajectories as a closed-form law, pointing toward learning interaction laws whose structure is not known a priori.

cs.LG · physics.atom-ph · physics.comp-ph

版本更新 · v2arXiv:2609.24954

JAREX: An Acquisition Function for Multi-Objective Algorithmic Process Characterization

Xinyang Li, Kevin Stone, Ajit Vikram

原文摘要

Pharmaceutical process characterization is central to Quality by Design because it defines how variations in process parameters affect the ability to meet product quality specifications, thereby supporting proven acceptable ranges and robust manufacturing. In practice, however, characterization still relies largely on factorial design of experiments (DOE) approaches, which are inefficient for resolving multivariate pass/fail boundaries in higher-dimensional spaces. While Bayesian optimization has transformed process optimization, adaptive methods for multi-objective process characterization remain lacking. Here, we introduce JAREX (Joint Acceptable Region EXploration), a Bayesian active-learning acquisition function for multi-objective process characterization. JAREX formulates characterization as a joint boundary-learning problem and adaptively selects experiments to recover the joint pass region defined by simultaneous satisfaction of threshold criteria across multiple objectives. JAREX combines an optimistic joint-feasibility mask with a multi-objective extension of randomized straddle, focusing sampling on the joint edge of failure. Our benchmark study suggests that JAREX provides more accurate and sample-efficient recovery of the joint pass region than factorial DOE, space-filling designs, and greedy objective-wise strategies over the full experimental budget range. For batched experimentation, it reduces the number of iterative process characterization experiments by more than half while preserving high accuracy for the boundary-identification task. Implemented in the open-source obsidian package, JAREX provides a modular framework for adaptive, data-efficient multi-objective algorithmic process characterization, supporting sample-efficient range finding in high-dimensional spaces.

stat.ML · cs.LG

版本更新 · v2arXiv:2610.09119

Which Buildings Are Artificial Intelligence-Ready? A Measurement-Based Assessment Framework for AI Question Answering and Actuation

Wooyoung Jung

原文摘要

Agentic artificial intelligence (AI) systems are becoming the interface to buildings, answering questions and controlling operations, but a building's readiness for them has not been systematically assessed. This study proposes a framework to quantify it. First, a building's knowledge graph sets two ceilings. The answerable-readiness ceiling is the share of operational questions its data could answer, and the actuation-readiness ceiling is the share of control actions it exposes. Second, a reference AI agent's accuracy on a fixed set of these questions shows how much of the ceilings is realized. On a simulated office, the agent realizes 0.62 of a 0.64 answerable ceiling, so missing data, not the AI, limit readiness, except in naming a fault's cause. Across 37 public real-building graphs, the median answerable ceiling is 0.16, and in 15 of 45, unlinked sensors lower it. The framework turns "is this building AI-ready?" into an auditable, ranked retrofit question.

cs.AI

新论文 · v1arXiv:2611.14801

Distributionally Robust BEST-RQ For Speech Representation Learning

Muhammad Khan, Xiaodong Cui, Brian Kingsbury, George Saon

原文摘要

BEST-RQ, based on random projection and quantization, has been a well-established self-supervised learning approach to learn speech representations from large-scale unlabeled audio due to its high performance and efficiency. However, in BEST-RQ representations learned by random quantization can be sensitive to the choice of quantization targets. To address this issue, we propose a distributionally robust extension of BEST-RQ that seeks representations robust across multiple quantization configurations. Specifically, we define groups by codebooks and formulate the training as a group Distributionally Robust Optimization (DRO) problem. Rather than optimizing solely for the average loss across groups, our approach adaptively reweights groups to emphasize those performing poorly. Additionally, we regularize DRO with a Kullback-Leibler divergence to prevent excessive distribution deviation. We evaluate our approach on the Multilingual LibriSpeech dataset and show that the distributionally robust BEST-RQ achieves superior performance over the conventional uniform average across multiple codebooks.

cs.CL · eess.AS · math.OC

版本更新 · v3arXiv:2605.10664

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions

Diancheng Kang, Zheyuan Liu, Ningshan Ma, Yue Huang, Zhaoxuan Tan, Meng Jiang

原文摘要

Activation steering controls language model behavior by adding directions to internal representations at inference time, but standard residual-stream steering can fail in stateful dialogue. We identify KV-cache contamination as a key failure mode: steered token states are stored and repeatedly reused, turning a local perturbation into cumulative coherence degradation. To address this challenge, we propose Gated Cropped Attention-Delta steering (GCAD), which extracts steering signals from system-prompt contributions to self-attention and applies them with token-level gating. Across persona-steering experiments, GCAD preserves trait control while substantially improving long-horizon coherence. On the main multi-turn benchmark, GCAD improves average coherence drift from -18.6 to -1.9 and raises turn-10 trait expression from 78.0 to 93.1. These results suggest that activation steering becomes more reliable when interventions follow the prompt-mediated pathways that models already use for behavioral control.

cs.CL · cs.AI

新论文 · v1arXiv:2611.14798

Who (or what) are they prompting? How CS students conceptualize and use LLMs

Nikitha Donekal Chandrashekar, Margaret Ellis, Sehrish Basir Nizamani, Naren Ramakrishnan

原文摘要

Large language models (LLMs) are increasingly used by students in computing courses. Yet computing education research has not fully examined how teaching and inclusion of LLMs shape students' conceptualization and understanding of LLMs, and how these factors relate to their subsequent use. Although prior work has examined prompting strategies and AI literacy, the relationships among students' conceptualizations of LLMs, their understanding of LLM mechanisms, and their prompting behavior remain underexplored. We studied these relationships in a post CS1 course that introduces students to the various CS subfields. We adapted the course to teach and allow the usage of LLMs. We administered an end-of-course survey, using open-ended items to elicit students' representations of LLMs and their explanations of how LLMs work, alongside closed-response measures of prompting style and prompt count (N=283 students consented and provided codable responses). Using qualitative coding, we categorized students' understanding of how LLMs work into four ordinal knowledge levels based on the Structure of Observed Learning Outcomes (SOLO) taxonomy and coded their metaphoric conceptualizations along two dimensions: persona type and relational role. Our results show that students most often conceptualized LLMs as human assistants, and context-rich prompting was the preferred style. The study contributes a two-dimensional characterization of students' LLM metaphoric representation and exploratory analyses also identified associations among these constructs showing how understanding these systems relates to their usage.

cs.CY

新论文 · v1arXiv:2611.14796

Selecting and verifying load-bearing periodic cells for elastic-wave band gaps with numerical acceptance checks

Shaoliang Yang, Jun Wang

原文摘要

A lattice mount must be stiff enough to support equipment and block vibration over a specified frequency range. Fast simulations help rank designs, but their errors can lead to a wrong acceptance decision. We present AMM (AgenticMetamaterials), an open-source Python tool that searches repeating cubic cells for targets on stiffness, the fraction of the cell that is solid, and band gaps. A complete band gap is an interval with no elastic-wave propagation in any direction. A separate checker allows for changes as the simulation grid gets finer. For gaps, it also tests for modes at sampled wave directions and wavelengths, so it can miss modes between samples. In 1,500 near-threshold tests of stiffness and solid fraction, the checker accepted 454 of 750 claims supported by the reference and none of the 750 false claims. Using the finest grid estimate alone accepted 42 false claims. In an independent finite-element comparison on ten new cells built from cubes joined by bars, the checker accepted one of two true gap claims and left all 17 false claims undecided. The reference could not classify the remaining claim. The protocol required every reported edge to err on the side of a narrower gap. The checker failed this test when it placed one upper edge too high, even though it left that thin-bar cell's claim undecided. Separating search from acceptance shows which targets the evidence supports and where a calculation needs further refinement.

cs.CE · physics.comp-ph

新论文 · v1arXiv:2611.14795

Model Checking Probabilistic Multiparty Session Types

Aleksander Junge, Leo Seojun Lee, Joseph Paulus, Nobuko Yoshida, David Parker

原文摘要

Distributed systems increasingly rely on randomisation; yet, verification techniques for such systems remain separate from the programming abstractions used to build them. This paper proposes the first probabilistic extension of general multiparty session types - typing sessions bottom-up, with no global specification - deriving quantitative properties of a multiparty session from its session types alone. Our types combine probabilistic choice with nondeterministic choice that abstracts data-dependent control, and allow separated choices, where each branch of a selection or branching may involve a different participant - a combination beyond previous probabilistic multiparty session type systems. To achieve this, we develop a novel probabilistic subtyping relation and merge operator, using which we prove type soundness and session fidelity. The semantics of typing contexts separates nondeterministic resolution, probabilistic choice, and synchronisation, so that every typing context induces a finite Markov decision process. On it, we derive sound lower bounds, holding under every fair adversarial scheduler, on the probability that a session is deadlock-free and live - the first behavioural property guarantees, partial or otherwise, for a probabilistic session calculus. To connect session types with probabilistic model checking, we introduce the first translation from session types into the PRISM modelling language, and prove it sound and complete with respect to our operational semantics. We implement the framework in the toolchain Prose, which translates probabilistic multiparty session typing contexts into PRISM and checks safety, deadlock-freedom, and liveness.

cs.PL

新论文 · v1arXiv:2611.14794

When Is Proof-Step Verification Worth Its Cost? Operating Requirements of Certify-or-Decline Verification for Open-Weight Mathematical Reasoning

Abhinav Gupta, Ben Slivinski, Yashica Patodia, Michael Saldivar

原文摘要

Proof-step verification rewrites a solver's answer into typed steps and uses LLM judges to certify or decline. We evaluate this pipeline on open-weight models across GPQA, AIME-2025, and a MATH-500 subset. Certification raises answer-key-match precision by 31 percentage points on the strong GPQA configuration and 46 on AIME, while answering only 35% and 21% of attempts. On near-ceiling MATH-500, gains are at most 5 points at a loss of 13-27 points of coverage. An output-token-matched self-consistency selector obtains 82.9% precision at 35% coverage, compared with the verifier's pooled 87.6% (paired difference: +4.8 points, 95% interval [-8.2, 18.5]). A diagnostic sample reveals a distinct failure mode: 52 of 88 completed problems end in invalid proof construction without ever reaching a judge. Finally, external-LLM acceptance of rendered certified chains is 61.0-67.6%, compared with 87.6% answer-key match. These audits lack human or formal-kernel ground truth. Open-weight auditor pairs show near-zero chance-corrected agreement, which measures consistency rather than accuracy against proof truth. These results motivate separate reporting of execution failures, selected-answer precision, and independently assessed proof validity. We provide per-configuration results and evaluation details in the supplement.

cs.AI · cs.CL · cs.LG

版本更新 · v4arXiv:2106.06801

Contrastive Semi-Supervised Segmentation with Class-wise Patches and Silhouette-Adaptive Consistency

Prashant Pandey

原文摘要

We present CSSL, a contrastive semi-supervised method for medical image segmentation. CSSL samples class-wise patches from labelled images and from confident pseudo-labels, separates their embeddings with a supervised contrastive loss with mixed hard negatives, and uses the Silhouette coefficient of the embedding clusters to set the strength of a feature-perturbation consistency loss. On MoNuSeg, patch embeddings are obtained by masked pooling of the encoder features of the unmasked image, nucleus rims form an additional contrastive class and a memory bank stores past embeddings; on CHAOS, patches are pixel-masked and no memory bank is used. The CSSL losses are added to a weak-to-strong pseudo-labelling recipe. We follow the protocols of published baselines: Self-Loop on MoNuSeg and DAFNet on CHAOS. On MoNuSeg with 50% labels, CSSL reaches 80.87 $\pm$ 0.17 nuclei F1 over three label splits, compared with 79.10 for Self-Loop and 78.73 for the same recipe without CSSL. On CHAOS T2 with 25% labels, CSSL reaches 89.01 Dice, compared with 82.0 for DAFNet and 88.94 for the recipe alone. Ablations show that the contrastive branch and, on MoNuSeg, the boundary class and feature-level pooling account for most of the improvement. Code: https://github.com/prinshul/CSSL

cs.CV

新论文 · v1arXiv:2611.14786

Source Assignment: Hardness, Approximation, and Parameterized Algorithms

Ruxandra Marinescu-Ghemeci, Adrian Miclăuş, Alexandru Popa

原文摘要

In Source Assignment, each pointer represents a substring of a common reference string and may have several possible source occurrences. The goal is to choose one source for each pointer while minimizing the number of distinct reference boundaries used by the selected sources. The problem was introduced by Qiu and Kingsford (Bioinformatics 2021) in the context of compact genome-graph construction. We distinguish \emph{Explicit Source Assignment}, where admissible source sets are given explicitly, from \emph{Complete Source Assignment}, where all occurrences of each pointer string are available. We show that Explicit Source Assignment has an exact value-preserving formulation as cardinality Set Cover with Pairs. For every fixed $\varepsilon>0$, it cannot be approximated within $(1-\varepsilon)\ln p$ unless $\mathrm{P}=\mathrm{NP}$, even over a binary alphabet when all pointers have length one, while known algorithms for Minimum-Weight Multicolored Subgraph give an $O(\sqrt{q\log p})$-approximation for arbitrary pointer lengths. Our main hardness result proves that Complete Source Assignment is APX-hard even over a binary alphabet, with equal pointer lengths and at most two sources per pointer. For edge-disjoint equal-length instances, we give a $3/2$-approximation and a $1.12585$-approximation when each pointer has at most two sources. We also introduce the rainbow-path length $λ$: the problem is polynomial-time solvable for $λ\le2$, and for $λ\le3$ we obtain a $(9/8+\varepsilon)$-approximation. We further give a polynomial-time test for deciding whether abstract source sets can be realized exactly as complete occurrence sets over an unrestricted alphabet. Finally, we give exact fixed-parameter algorithms for several structural parameters, and show that Complete Source Assignment is fixed-parameter tractable in the alphabet size and maximum pointer length.

cs.DS

版本更新 · v2arXiv:2510.19806

The Art of Asking: Multilingual Prompt Optimization for Synthetic Data

David Mora, Viraat Aryabumi, Wei-Yin Ko, Sara Hooker, Julia Kreutzer, Marzieh Fadaee

原文摘要

Synthetic data has become a cornerstone for scaling LLMs, yet its benefits for multilingual modeling remain bottlenecked by translation. Translation for data augmentation inherits English-centric framing and style and neglects cultural dimensions, ultimately constraining model generalization. We argue that the overlooked prompt space offers a more powerful lever for improving multilingual performance. We introduce a lightweight framework for prompt-space optimization, where translated prompts are systematically transformed for Naturalness, Cultural Adaptation, and Difficulty Enhancement. Using an off-the-shelf multilingual LLM, we apply these transformations to prompts from 12 languages spanning 7 families. Under identical data conditions, our approach achieves substantial and consistent downstream improvements over the translation-only baseline: +4.7% accuracy on Global-MMLU, +2.4% XCometXL on Flores and +35.3% wins in preferences on mArenaHard. These gains are further corroborated by human preferences on open-ended creative writing. We establish prompt-space optimization as a simple yet powerful paradigm for building multilingual LLMs that are more robust, culturally grounded, and globally capable.

cs.CL

新论文 · v1arXiv:2611.14783

A Multi-Prover Interactive Protocol for Certification of Resolution Refutations

Philipp Czerner, Adrian Krauss, Josia Pietsch, Javier Esparza

原文摘要

A resolution refutation of a SAT-formula $\varphi$ is an independently checkable certificate proving that $\varphi$ is unsatisfiable. However, the refutations of $\varphi$ can be exponentially larger than $\varphi$. So, a verifier needs worst-case exponential time in the size of $\varphi$ to check their correctness. We present ResCertify, an interactive probabilistic protocol in which the verifier interacts with multiple independent, untrusted provers (MIP) to check the refutation with very low probability of error. The verifier runs in linear time in the size of $\varphi$. The provers require time $\mathcal{O}(n \cdot |Φ|)$, where $Φ$ is the resolution refutation and $n$ is the number of variables of $\varphi$. If they can be trusted to act as a multilinear oracle on certain queries, a single prover suffices, and the runtime can be improved to $\mathcal{O}(|Φ|)$. Finally, while the focus of our work is theoretical, we also report on a proof-of-concept implementation of ResCertify.

cs.LO

新论文 · v1arXiv:2611.14782

Bayesian inference with a polymer probabilistic computer

Travis Wanless, Yuriy Bandera, Lubomir Kostal, Stephen H. Foulger

原文摘要

Bayesian inference weighs competing explanations of noisy data by sampling from a posterior distribution, a task that deterministic hardware performs only by simulating randomness. Probabilistic bits (p-bits) supply this randomness physically, but the few hardware networks of p-bits built so far use inorganic devices, address optimization or learning, and are judged without reference to the spread expected from finite sampling. Here we report a probabilistic computer built from solution-processed conjugated-polymer memristors. Four polymer p-bits, tracked online for drift and coupled in a closed Gibbs-sampling loop, sample competing, degenerate and cooperative posteriors over sixteen states with statistics that fall within the range of ideal p-bits, both in their distribution and in their autocorrelation, without any parameter fitted to the target. This agreement extends to new simulated sensor readings whose outcome was predicted before each run, including readings that point away from the true answer. Reprogrammed in software, the same circuit and controller reproduces the invertible AND gate of earlier p-bit networks, with its temperature fixed in advance rather than fitted. Polymer p-bits thus open a route to printed, large-area probabilistic computers that infer alongside their sensors.

cs.ET · cond-mat.dis-nn · cond-mat.mtrl-sci · cond-mat.soft

新论文 · v1arXiv:2611.14781

Robust Task-Aware State Estimation from Pre-trained Perception

Guy Azran, Sarah Keren

原文摘要

Agents that perform state estimation using pretrained perception models, e.g., Vision-Language Models (VLMs), face the fundamental challenge that predictions are uncertain. Still, most approaches determinize these predictions into a single best-guess state and make decisions as if it were ground truth. This is brittle because when the guess is wrong, the agent's decisions may be unsafe or infeasible. At the other extreme, maintaining a full probability distribution over symbolic states is often intractable. We propose a middle path, drawing on the underlying probabilities the perception model already produces (e.g., a VLM's output-layer token distributions) to construct a $θ$-confidence Minimal Belief Set (MBS$_θ$), a minimal subset of states whose cumulative probability meets a user-specified belief-mass threshold. The minimality of an MBS$_θ$ ensures that the agent considers only the most likely states, enhancing robustness in downstream decision-making while maintaining computational tractability. Using a novel three-layered belief system, we efficiently approximate individual state probabilities from per-feature predictions and task-level constraints. On this belief construction, we base our MBS$_θ$ construction, while maintaining only a lean factored representation of the belief. Our algorithm, Probability-Ranked Valid State Search (PRaVaSS), takes advantage of the unique structure of this belief system to construct an MBS$_θ$ with correctness guarantees and polynomial complexity in the unconstrained case. Using a VLM as the backbone of our state estimator, we validate our method's theoretical properties, scalability with problem size, and improvement of downstream decision-making when used in a task-planning pipeline.

cs.AI