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

+1085新增论文
+884其中 AI 相关
+130已有论文修订
+94新增 AI 导读

arXiv 最近完成扫描:2026-11-19 06:59 · 全站最近入库:2026-11-19 06:59各信源进度 →

统计口径与数据新鲜度

统计 2026-11-18 09:26 — 2026-11-19 09:26(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:2026-11-19 06:59最近发现新增版本:2026-11-19 06:59库内论文最新更新时间:2026-11-18 08:00

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。

26,871 已收录论文22,375 AI 相关8,156 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 325 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-19 06:57(北京时间)
版本更新 · v4arXiv:2611.12657

When Does Sliced Inference Pay Off? Resolution, Tiling and Dense Fusion for Small Object Detection

Stefan Becker, Jens Bayer, David Münch, Wolfgang Hübner

原文摘要

Sliced inference such as SAHI is the standard remedy for small objects in high-resolution imagery, but its gains are usually measured against low-resolution baselines. We revisit sliced inference with scale as the explicit variable and ask when tiling actually pays off. We compare nine dense detectors, eight YOLO models of four generations and FCOS, on VisDrone and DOTA, under pre-specified hypotheses, calibration-only tuning and a single evaluation on held-out test data. Two findings emerge. First, much of the benefit attributed to slicing is a resolution effect: on VisDrone, a single pass trained at a higher resolution outperforms tiling with box fusion (merging decoded boxes) for every detector that could be trained at that resolution, and at equal compute a higher resolution helps more than a larger model. Second, tiles carry complementary information that box fusion wastes. We introduce SCAHI (Slicing with Coverage-Aware Hyper Inference), which accumulates the dense outputs of all passes with weights that give every image position the same total coverage, and reads objects off a peak grid fine enough to keep crowded neighbours apart. On identical passes, SCAHI outperforms box fusion for every YOLO detector on both datasets and matches it for FCOS. On DOTA, whose images far exceed the resolution a single pass can be trained at, tiles with SCAHI beat the best single pass for every detector, by 4.3 to 5.6 mAP50 at 2048 px, and replacing only SAHI's merge step by SCAHI adds 2.3 to 4.0 mAP50. An ablation attributes the gain to the border weighting and the resolution of the peak grid. Source code is available at https://github.com/beckerio/scahi.

cs.CV

新论文 · v1arXiv:2611.20484

Agent-Guided Configuration Tuning for SMT Solvers

Cong Wang, Qiaoyi Zheng, Zhaoguo Wang, Fan Long

原文摘要

SMT solvers expose many configuration choices that determine how well their existing algorithms perform on a workload. We present, to our knowledge, the first workflow that uses general-purpose LLM agents to tune SMT solver configurations. Agents read solver documentation, source code, and persistent experiment records to propose option changes, and a deterministic harness measures each proposal, validates the returned models, and selects changes separately for each competition Division. Over roughly nine weeks, our tuning campaign recorded 4,545 proposals for cvc5 across 13 Divisions and 36 logics. We evaluate the tuned configurations on the 2026 and 2025 SMT-COMP ModelValidation benchmarks under a 24-second cutoff. On problems not used for training, they validate more problems than official cvc5 in ten of the 13 Divisions in 2026 and in eight in 2025, with geometric-mean gains of 7.8% and 6.8% across Divisions. On the complete official selections, which include training problems, the tuned floating-point counts exceed those of every individual competition participant in both years. Further studies show how configuration changes make existing algorithms effective by changing access to strategies, scheduling, and option compatibility, and how benchmark families trade off against one another. The agents' experiment traces contain hypothesis-based proposals in every tuning group, and 64.3% of the extracted hypotheses draw on prior measurements. These results show that agent-guided tuning can specialize a mature SMT solver and that accumulated experimental evidence informs sustained search.

cs.AI

新论文 · v1arXiv:2611.20435

iEEG Can Help EEG: Improving Clinical Scalp EEG Analysis through Intracranial Priors

Xianglong Bao, Yi Ding, Zengxiang Li, Nan Liu, Kewen Wang, Zhe Wang, Ming Jin

原文摘要

Scalp electroencephalography (EEG) is a cornerstone of clinical brain monitoring, making the quality of its signal fundamental to downstream tasks and models. However, that signal is heavily contaminated by non-neural interference and blurred by volume conduction. Existing methods operate almost exclusively within the scalp modality, so they inherit the contamination they would need to overcome. In practice, intracranial EEG (iEEG) observes the neural activity with far higher fidelity, yet it is invasive and therefore scarce, and the two modalities are rarely recorded as pairs at scale, share no common channel space, and offer no clean ground truth to directly supervise a mapping. Motivated by the neuronal synchronization mechanism shared by the two modalities, we propose a reprogramming approach that fits a signal-to-signal model on iEEG alone and freezes it as a cross-modal prior. We introduce ReBrain, a framework to transfer an iEEG-learned prior to scalp EEG, which re-expresses any scalp signal through that prior and can be inserted upstream of any downstream model with no pairing or fine-tuning. ReBrain consistently improves host models across recordings from multiple hospitals and across architecture families, and our analysis traces this gain to genuine neural structure carried by iEEG rather than to the architecture itself. This work provides evidence that a scarce clean modality can serve as a transferable prior for an abundant noisy one, offering a general recipe for cross-modal transfer among biosignals. Our code is available at https://github.com/Xianglong-Bao/ReBrain.

cs.LG · eess.SP · q-bio.NC

版本更新 · v2arXiv:2609.36724

Routing in Gradient Space: Balanced Usage Is Not Expert Specialization

Yuchen Li, Mingyu Du, Zongqi Fan, Nguyen H. Tran, Ken-Tye Yong

原文摘要

Sparse expert models can distribute traffic evenly while still grouping incompatible training signals within the same experts. We study routing as a gradient-partitioning problem and introduce gradient-aligned routing (GAR), whose load-normalized router objective rewards grouping observations with aligned gradients. On five multi-task text-classification mixtures, we compare GAR with task-loss-only routing, gradient-combination and gradient-conflict methods, and load-balancing losses. With a fully trainable RoBERTa backbone and classification-head experts, GAR has the highest aggregate validation accuracy, 1.07 percentage points above task-loss-only routing. With frozen DeBERTa and Qwen3-1.7B backbones and low-rank adapter experts, it again ranks first, 1.10 points above task-loss-only routing, with better-balanced expert load and higher gradient-mass purity, the share of each expert's gradient-norm mass from its dominant task; the load-balancing losses flatten load further but leave this purity near its task-loss-only level. Top-1 routing, trainable full-parameter feed-forward network (FFN) experts, and a larger backbone also show positive aggregate gains. In component ablations, the objective's cross-observation alignment terms alone retain GAR's accuracy, and GAR keeps its accuracy when observations are formed without task identity; supervising the router with task labels raises task-expert association without adding accuracy. The results distinguish expert-load balance from gradient-based routing organization and show the predictive value of gradient-informed routing in multi-task text classification.

cs.LG

版本更新 · v2arXiv:2605.17671

PEIRA: Learning Predictive Encoders through Inter-View Regressor Alignment

Michael Arbel, Basile Terver, Jean Ponce

原文摘要

Non-contrastive self-supervised learning (SSL) is an effective framework for predictive representation learning, but popular (and in practice effective) methods such as SimSiam, BYOL, I-JEPA or DINO, which rely on a form of self-distillation to train a teacher-student network, remain poorly understood as they typically do not minimize a well-defined objective. We analyze the dynamics of a variant of the Joint Embedding Predictive Architecture (JEPA) using a regularized linear regressor to predict the learned representations of two views of the data from one another, and fully characterize its stability: non-collapsed stable equilibria align with leading nonlinear canonical correlation subspaces, while collapsed equilibria may also be stable attractors. Motivated by this result, we introduce PEIRA, a non-contrastive SSL method with an explicit objective defined through the trace of the optimal linear regressor. We show that its only stable equilibria are nontrivial global minimizers and recover the same canonical correlation subspaces, with regularization selecting the effective dimension. Experiments on ImageNet-1K and CIFAR-10 show PEIRA is competitive with VICReg and LeJEPA baselines, and qualitative empirical results support the theory. Project page: https://peira-ssl.github.io. Code: https://github.com/MichaelArbel/peira-loss (loss implementation), https://github.com/Basile-Terv/PEIRA (experiment reproduction).

cs.LG · cs.AI

新论文 · v1arXiv:2611.20314

NF-VLA: A Normalizing-Flow Baseline for Vision-Language-Action Learning

Ruichen Deng, Qinyu Zhao, Yong-Hyun Park, Mutian Tong, Liang Zheng, Jiatao Gu

原文摘要

Autoregressive normalizing flows (NFs) provide a tractable density for learning and sampling continuous robot actions. We show that a single maximum-likelihood objective can train vision--language--action (VLA) policies competitive with leading methods on manipulation benchmarks. We present NF-VLA, which combines conditional autoregressive normalizing flows with score-based action refinement. Its deep--shallow architecture concentrates visual, language, and proprioceptive prefix conditioning in a high-capacity deep block, while unconditioned shallow blocks provide a shared invertible action transformation. The vision--language backbone and flow are jointly trained by exact maximum likelihood on Gaussian-smoothed action chunks. The same learned density supplies a single score-based correction at inference, refining generated actions to account for training-time smoothing without a separately trained denoiser. Trained on LIBERO from a pretrained vision--language backbone without robot pretraining, NF-VLA achieves 98.35\% success across the four standard suites and 72.6\% on LIBERO-Plus using the same checkpoint without further fine-tuning. Ablations show that deep-only prefix conditioning and score-based refinement both improve closed-loop success. Together, these results establish carefully structured autoregressive flows as a competitive approach to VLA control, with one learned density supporting policy training, action generation, and refinement.

cs.RO · cs.AI

新论文 · v1arXiv:2611.20289

Benchmarking LLM Inference on Desktop-Scale NVIDIA Grace Blackwell Clusters

Kevin Bhimani, Christian Escobar, Daniel Mihovch, Rudolf Eigenmann, Sunita Chandrasekaran

原文摘要

Two NVIDIA DGX Spark GB10 systems connect directly over 200 GbE to serve larger language models and increase inference capacity, but the benefits depend on how they are deployed. We benchmark 17 open-weight checkpoints spanning dense and mixture-of-experts architectures, 8B to 235B parameters and four weight precisions on one and two RDMA-connected systems. Using vLLM, we compare tensor parallelism (TP=2), pipeline parallelism (PP=2) and a two-replica baseline at concurrency 1 to 64, with three independent launches per configuration. TP=2 delivers a median 1.59x peak-throughput gain over one Spark, versus 1.39x for PP=2 and has higher throughput in 93 of 96 matched comparisons with separated 95% Welch intervals. It reduces unloaded inter-token latency to 0.52x to 0.73x of one Spark, whereas PP=2 provides little reduction. TP=2 also achieves the highest goodput under a strict 100 ms inter-token latency objective wherever both strategies are supported and the objective is met. For models fitting on one Spark, TP=2 generally favors low-concurrency serving, while replication offers higher throughput at high concurrency. One Spark uses less GPU energy per output token at matched concurrency. Across three precision ladders, NVFP4 reduces unloaded latency more than adding a second Spark at BF16. NCCL microbenchmarks and comparisons with an earlier TCP campaign suggest that collective communication overhead contributes to the advantage of TP=2. These results provide deployment guidance for desktop-scale inference.

cs.DC · cs.AI

新论文 · v1arXiv:2611.20230

A Matrix with Steps: An Episodic and Procedural Memory Method for Tool-Using Agents and Robots

Jürgen Paulhart, Peter Schüller

原文摘要

Agents that call tools and robots that execute tasks are challenged by two shortcomings: they repeat what failed before, and they learn from runs that they only believe succeeded. We describe the construction of a memory that addresses both, built for a production agent platform. Every run is recorded as an episode, and only an outcome confirmed by an external check, a test, a tool result or a person counts; the agent's own claim never does. What is learned from episodes is kept in a memory with two parts: an associative matrix, written with the delta rule, which finds the procedure that applies to a situation; and a section of ordered, addressable step records that answers what comes after a given step, where a step is a goal of the agent, not a motion. Because the matrix sums its associations, it keeps no order and cannot return or delete a single one; it serves as the finder, and order, outcome counts and deletion live in the records. We describe how execution outcomes are stored, how episodes are recalled and used for solving tasks, and the file format that stores the memory. In simulation, a robot that judged its own success and stopped when it believed it was done reported $96\%$ of its tasks as done and solved $35\%$; with external verification it solved $79%$. In a store of $100000$ synthetic episodes ($457$MB on disk), the oldest episode was still found first, no episode from another person's memory was returned, and a success vouched for only by the model was never recalled; a lookup took $16$ms on a laptop and $40$ms on a server node (median). Task outcomes were measured up to about $10000$ confirmed records, where the exact records solved every revisit at the first attempt.

cs.AI

版本更新 · v3arXiv:2607.11250

Do Multi-Agent LLMs Explore Each Other?

Hyeong Kyu Choi, Jiatong Li, Wendi Li, Xin Eric Wang, Sharon Li

原文摘要

Exploration is essential for reliable autonomy in multi-agent systems, yet it remains unclear whether large language model (LLM) agents can explore effectively when interacting with one another. We show that modern LLM agents often fail to do so, exhibiting myopic and polarized interaction patterns that lead to suboptimal coordination and increased regret. We formalize this challenge as the Multi-Agent Exploration problem, modeling it as a partially observable stochastic game (POSG), in which agents must probe peers to infer their capabilities and identify effective interaction strategies. To address this, we introduce Multi-Agent Contextual Exploration (MACE), a lightweight framework that explicitly promotes exploration through structured peer selection. Across both contextual and parametric diversity settings, exploration behavior and downstream task performance substantially improve when MACE is applied. We further provide theoretical insight that MACE achieves a sublinear expected-regret bound. Overall, our results highlight a fundamental limitation of current LLM agents and underscore the importance of explicitly guided exploration for reliable multi-agent autonomy.

cs.MA · cs.AI

新论文 · v1arXiv:2611.20126

A Corpus-Driven SentencePiece Unigram Tokenizer for Sindhi: Design and Benchmark Against Fourteen Existing Tokenizers

Ali Nawaz, Muqadas Fareed, Soonh Sadaf

原文摘要

Sindhi is a low-resource Indo-Aryan language written mainly in an extended Perso-Arabic script. Several Sindhi tokenizers have been released on public model hubs, but they have not been compared with one another on shared data. We train SentencePiece Unigram tokenizers with byte fallback at four vocabulary sizes (32k to 150k) on about 1.10 million Sindhi sentences and a smaller, repeated Urdu, Persian and Arabic slice, from four public sources, and compare them with fourteen existing tokenizers on four evaluation sets: FLORES+, Sindhi Wikipedia, a news-derived sentiment corpus, and a contemporary news corpus. The metrics are fertility, single-token retention, UNK rate, byte-fallback rate, and round-trip character loss. At 64k, our tokenizer averages 1.147 tokens per word, against 1.208 for the best existing tokenizer and 1.512 for XLM-RoBERTa, and keeps 88.6% of words as single tokens. On fertility, single-token retention and UNK rate it is best or tied for best in all four evaluation sets at 64k and above; at 32k it trails one baseline on FLORES+. Margins over the best Sindhi-adapted baselines are small (0.002 fertility on FLORES+ at 64k), and we report no significance tests. A round-trip check finds no character loss for our tokenizer and silent character deletion (up to 0.85% of characters) in two existing ones; seven others lose less than 0.1%. We evaluate intrinsic metrics only, so effects on downstream tasks remain untested. Code and tokenizer files are public (DOI: 10.5281/zenodo.23962474).

cs.CL

版本更新 · v4arXiv:2611.10292

Continual Instruction Tuning for LLM-based Management of Continuum Resources

Lanpei Li, Antonino Vaccarella, Vincenzo Lomonaco, Massimo Coppola

原文摘要

Large language models (LLMs) are increasingly explored for resource management across the Cloud-Edge continuum. However, embedding LLMs directly inside latency-sensitive control loops places inference latency, execution jitter, and hallucination risks on the critical path. This study argues for a different use of LLMs: a continually instruction-tuned intent compiler that runs once per management intent or declared context change, above a fast, deterministic tactical controller that handles all low-latency runtime actions. We formulate intent compilation as a continual instruction tuning problem, because management intents arrive sequentially and accumulate while infrastructure and workload contexts drift. Unlike recent LLM-based management frameworks, our design scores the model's outputs by execution, so that catastrophic forgetting can be measured in the executed objective of the underlying controller. On a continuum simulator with a 2B backbone and parameter-efficient adapters, preliminary results show that naive sequential tuning forgets earlier intents in the executed objective, that replay of stored examples nearly prevents this forgetting, and that the tuned compact compiler performs on par with simpler translators while being more robust than a grammar parser to rephrased intents. Code and data are available at https://github.com/lilanpei/continuum-intent-compiler.

cs.DC · cs.LG

新论文 · v1arXiv:2611.19793

IANN: A Unified Framework for Equivariant InterAtomic Neural Networks with User-Friendly Implementation

Changzhi Ai, Xueping Qin, Dimosthenis Sokaras, Frank Abild-Pedersen, Johannes Voss

原文摘要

Machine-learning interatomic potentials (MLIPs) now approach density functional theory (DFT) accuracy at a fraction of its cost, but the practical benefit is limited by software fragmentation: each state-of-the-art architecture is released as a separate codebase, with its own data format, configuration system, training loop and molecular-dynamics interface. Comparing two architectures on the same dataset, or moving a trained model into a production simulation, therefore requires substantial re-engineering. We introduce IANN (InterAtomic Neural Network), an open-source PyTorch framework that reimplements equivariant graph neural networks - PaiNN, NequIP, MACE, Allegro, EquiformerV2, EquiformerV3 and UMA - behind a unified data object, a unified trainer and a single deployment route, so that the architecture is a one-word change in a configuration dictionary. Forces, stresses and virials are obtained for every architecture from one analytic differentiation of the energy with respect to the edge vectors of the atomic graph, exact under periodic boundary conditions. Distributed training is configured automatically from the SLURM, PBS or MPI environment, so the same script runs unchanged from a workstation to a multi-node GPU allocation. Self-describing checkpoints serve directly as ASE calculators, as ensembles reporting uncertainties, or as TorchScript models run inside LAMMPS through bundled C++ pair styles with single- and multi-GPU inference. Trained on a common dataset of millions of structures, the architectures can be compared under identical conditions. We further release PaiNN foundation models trained on 41.9 million curated DFT structures from MPtrj, Alexandria, OMat24, MatPES, OC20, OC22, OC25, MPr2SCAN and MatPESr2SCAN, spanning the PBE, RPBE and r2SCAN functionals. Additionally, IANN provides an agent interface to facilitate integration with autonomous agents.

cond-mat.mtrl-sci · cs.LG · physics.comp-ph

新论文 · v1arXiv:2611.19700

Where Should LLM Agents Adapt at Test Time? In Context, in Weights, or Both

Parsa Mirtaheri, Mohammadreza Armandpour, Keivan Alizadeh, Iman Mirzadeh, Parshin Shojaee, Aryo Lotfi, David Harrison, Mehrdad Farajtabar

原文摘要

LLM agents operating repeatedly in the same environment, such as a code repository, a database, or an application suite, encounter tasks that share latent structure. Experience can reveal this structure, allowing agents to reuse environment knowledge and learn from environmental feedback. Agents can adapt at test-time using this experience through two channels: retaining it in context and updating their weights. These channels are usually studied separately, leaving unclear what each contributes and when they complement each other. We compare in-context learning (ICL), reinforcement learning with verifiable rewards using GRPO, and self-distillation using SDPO, alone and in combination, using a shared online protocol and modular framework. Agents encounter matched task streams, are evaluated on each task before adapting to it, and learn only from their own attempts and environment feedback. We evaluate 14 adaptation mechanisms and a stateless baseline with Qwen3.5-9B on 109 tasks from four environments drawn from SWE-bench Verified, AppWorld, ScienceWorld, and BIRD-Critic-SQLite, using five task orders per environment. ICL provides the largest standalone gain: its best policy raises mean success across environments from 28.8% to 49.4%. Adding GRPO to this policy achieves the highest mean success, 51.6%. While adding GRPO improves the overall mean under every ICL policy, SDPO has mixed effects. Trace analyses show fewer lookups and shorter attempts under ICL, consistent with reusing knowledge of the shared structure. However, ICL also frequently skips inspection and verification for the current task. Adding GRPO to ICL helps restore these behaviors while preserving its reduced environment lookups. These observations suggest that reinforcement learning with verifiable rewards can refine how agents use retained experience, motivating the joint study of context management and weight updates.

cs.LG · cs.AI · cs.CL

版本更新 · v2arXiv:2607.22450

A Maximum Entropy Implementation of Differential Privacy Under Linear Invariants

Ryan Lafferty, Anindya Roy

原文摘要

Differential privacy is the standard for ensuring data privacy and is widely used in major data publications, including reporting results from the U.S. decennial census. Common implementation of differential privacy uses independent Gaussian or Laplace noise addition to the database. However, there could be aggregate (linear) queries to the database that are excluded from the privacy budget, for example, state totals that can not be perturbed due to constitutional mandates. Any implementation of a differential privacy is required to honor these constraints, also referred to as invariants. Under aggregation constraints, the noise vector is no longer independent and the traditional differential privacy guarantees have to be re-evaluated. We propose a high entropy differential privacy implementation that maintains the aggregation invariants with probability one or exponentially close to one and derive the privacy guarantees for the implementation under the invariants. The theoretical proof also gives a solution to an open question about the null space of correlation matrices. Moreover, the methodology has general use in the context of sampling from normal mixture models under linear equality constraints.

cs.CR · stat.ME

版本更新 · v2arXiv:2610.39418

LinQuant: Functional Invariances for Low-Precision Linear Recurrences

Neehal Tumma, Philipp Nazari, Rohan Cherukuri, Daniela Rus, T. Konstantin Rusch

原文摘要

Linear recurrences are emerging as practical alternatives to softmax attention in large language models, appearing in both standalone and hybrid architectures. However, despite their rapid emergence, low-precision inference remains much less developed for these models than for softmax attention. During prefill, chunkwise-parallel algorithms cast much of the recurrence as dense matrix multiplications, suggesting a natural path to low-precision execution. Simply quantizing these operations in their native form, however, can be fragile. In this work, we identify functional invariances that preserve the full-precision computation while making the operands more amenable to quantization. Building on this observation, we introduce LinQuant, a framework for low-precision recurrent prefill based on seven such invariances, including rotations, diagonal rescalings, and alternative placements of recurrent decay and write factors. We derive when these transformations remain valid across common recurrence families spanning linear, gated, and delta-rule updates, and show that they provide complementary degrees of freedom for improving low-precision accuracy. We implement these transformations in INT8 and INT4 kernels and evaluate them on the recurrences of Nemotron-H, Qwen3.5, and Kimi Linear, recovering most of the accuracy lost to quantization while accelerating recurrent prefill. We also augment SSDi8, an existing low-precision prefill pipeline for Mamba-2, improving perplexity at matched prefill latency on pretrained Mamba-2 and Nemotron-H models. Altogether, these results demonstrate the practical utility of invariance-based quantization for recurrent models.

cs.LG

新论文 · v1arXiv:2611.19587

Explicit Regularization for Flow Matching

Lars Ruthotto, Eran Treister

原文摘要

Learning the underlying probability distribution from samples is a core problem in statistics and applied mathematics. Generative models based on diffusion and flow matching learn a velocity field whose flow transports a known distribution to the data distribution. They produce novel, realistic samples, although the empirical learning problem is ill posed near the data distribution. Its minimizer transports the samples onto the training data, and small perturbations of the data change the velocity field strongly near the end of the flow. The observed generalization is often attributed to implicit regularization by the network parameterization, early stopping of the training, and the time integration during sampling. Explicit regularization has advanced the theory and computation of many inverse problems and has recently been proposed to prevent memorization in generative models. With a regularizer on the magnitude of the velocity, however, the field approaches the memorizing velocity unless the regularization parameter grows to infinity fast enough, and it then stays bounded by a constant that is small only for large parameters. We show that penalizing the derivatives of order $k$ of the velocity renders the $D$-dimensional empirical problem well posed uniformly in time if $2k > D$ and the training points are in general position. Our experiments in one and two dimensions discretize the velocity with splines and use direct solvers for the regularized least-squares problems to exclude optimization error. Refining the discretization in space and time changes the error of the Hessian-regularized flows only marginally, whereas without the regularizer the error tends to that of memorization. These flows generalize, with errors comparable to or lower than those of well-tuned kernel density estimates.

cs.LG · math.NA

版本更新 · v2arXiv:2611.14357

The Unused Input: The Initial State as a Conditioning Interface for Recurrent TSFMs

Bernhard Voggenberger, Christian Ganhör, Levente Zólyomi, Sebastian Böck, Sepp Hochreiter

原文摘要

Time series foundation models (TSFMs) forecast unseen series without task-specific training. Increasingly they are built on recurrent backbones, whose constant per-step cost makes long-context inference cheap, and which, unlike transformers, carry a state. A recurrent backbone computes its prediction from two inputs: the observed context window and the initial recurrent state. In practice only the first is utilised and the initial state is unused by convention. We ask whether the initial state can be utilised in recurrent TSFMs. Using the xLSTM-based zero-shot forecaster TiRex as a case study, we freeze the backbone and train only a small head that writes a context vector into the initial state, so that any change in accuracy is attributable to the injected state rather than to adapted weights. A spectral summary of the input window, learned once and applied unchanged to every series, lowers mean GiftEval MASE by 0.175%, with every seed improving, which neither a parameter-matched LoRA adapter nor a learnable static state reproduces. An embedding of text attached to each series, evaluated on six MoTime datasets, reveals two roles for the state. As a per-dataset adaptation surface, a static state lowers MASE by 0.04-6.96% of the zero-shot baseline (8-70% of the gain of full fine-tuning). As a channel for exogenous information, the text-conditioned state lowers it by a further 0.15-0.41% of the zero-shot baseline on four of six datasets and, combined with a LoRA adapter, exceeds full fine-tuning, which has no access to the text, on one of them. In TiRex, the effect concentrates in the mLSTM blocks and is strongest at short context. The results are preliminary and the improvements small, but they point to the initial state as a direction for further research.

cs.LG

新论文 · v1arXiv:2611.19295

Optional Traps: An Empirical Study of LLM Agents Against Credential Deception and Alert-Driven Revocation

Raditio Ghifiardi

原文摘要

Deception (honeypots, decoys and canary credentials) rests on a simple premise: legitimate users never touch a fake asset, so any touch is a high-confidence signal. We ask whether that premise holds when the attacker is a large language model (LLM) agent, and what deception contributes once ordinary hardening is in place. We built an isolated, hash-chained lab around a mock Records API and report four studies. (1) In an exploratory run, a local 9.7B-parameter model ignored every optional decoy, read a genuinely leaked service key from a debug endpoint and exported the protected records in both static and adaptive deception modes. (2) After removing the leak, planting a canary credential in its place and revoking the offending client, a frozen 60-session study with a 4B-parameter model found export success of 20/20 at baseline, 0/20 with hardening alone and 0/20 with hardening plus deception (exact paired McNemar p = 1.9e-6 baseline vs. defended; p = 1 hardened vs. defended). Deception added detection and per-client revocation in 19 of 20 defended sessions, but no measurable prevention beyond hardening. (3) On Google Cloud, a real OpenCanary login alert was authenticated, signed and turned into revocation of the attacking client; the first denied export followed within an observed upper bound of 303 ms, a separate legitimate client completed 100/100 requests, and all 14 control checks passed. (4) In multi-agent swarm engagements, agents that used the canary were revoked 39-59 s after the bait was served without affecting siblings, but one model-driven legitimate agent also took the bait, a false-positive risk specific to LLM clients. Against LLM agents, traps must sit on the attacker's path rather than beside it, closing real weaknesses does the preventive work, and deception's measurable value is fast, attributable detection. All results are lab-scale and are not a production claim.

cs.CR

版本更新 · v2arXiv:2606.16613

CoffeeBench: Benchmarking Long-Horizon LLM Agents in Heterogeneous Multi-Agent Economies

Issa Sugiura, Daichi Hattori, Kazuo Araragi, Keita Ogawa, Shota Onose, Taro Makino, Teppei Usuki, Takashi Ishida

原文摘要

LLM agents are increasingly capable of handling complex, long-horizon tasks and may eventually take on business management, a role traditionally performed by humans. Deploying them safely requires monitoring their capabilities and behavior in environments that reflect real economies. We introduce CoffeeBench, a benchmark built on a compact simulation of a coffee supply chain with realistic business mechanics. Six autonomous LLM coding agents run its farmers, roasters, and retailers, and negotiate and trade with one another to maximize their own net income over 90 simulated days. We evaluate eight recent proprietary and open-weight models by letting each run one roaster while a fixed background model runs the other five firms, and scoring that roaster's net income. Claude Sonnet 5.5 earns the highest net income, closely followed by GPT-6 Astra. The leaders sell at higher prices, and the best model also buys its inputs far more cheaply, whereas weaker models trade at thin margins. Since performance targets are a common cause of accounting fraud in human firms, we also test whether agents under KPI pressure inflate their numbers. Under a revenue target, circular trades occur for some pairs of models: one agent proposes that the two rival firms sell goods back and forth at inflated prices so that both book revenue, and the rival accepts. We hope CoffeeBench serves as an early step toward a future in which LLM agents manage firms without misconduct. We release our code and agent trajectories.

cs.AI

新论文 · v1arXiv:2611.18111

Multi-GPU Tridiagonal Partition Algorithm

Milena Veneva

原文摘要

This paper presents a multi-GPU version and a multi-GPU recursive version of the parallel partition method for systems of linear algebraic equations with tridiagonal coefficient matrices and their implementations in CUDA. Computational experiments on one and two GPU cards, for different system sizes and with different number of recursive steps are conducted. The results are analyzed and discussed.

cs.DC

版本更新 · v2arXiv:2609.35075

ReCo: When to Relocate Sensor Kits under Deployment Constraints -- A NILM Case Study

Haokun Chen, Yu Tong, Yehai Chen

原文摘要

Many sensing tasks obtain training labels only by deploying instruments in the field. With a limited number of sensor kits, a collection deadline, and measurement downtime at every move, the collector must repeatedly decide whether to stay at the current site or relocate. We study this decision in non-intrusive load monitoring (NILM), which estimates the power drawn by individual appliances from a home's main meter and is trained on data from homes temporarily fitted with appliance-level sub-meters. In NILM, appliance usage varies with the appliance, season and climate, and the value of new data depends on how diverse the combinations of target operation and background load are. To address this, we propose a constraint-based relocation framework and instantiate it for NILM as ReCo (Relocation by Coverage gain). ReCo counts new operating regimes in a joint target--background feature space, forecasts each home's future gain from the data collected so far, and each night weighs the gain of staying against the gain of moving elsewhere after the downtime. In replayed deployments on the Plegma dataset under two kit counts and two downtime costs, ReCo outperforms fixed-dwell and count-based schedules and a threshold rule using the same metric in every setting. Its advantage is not explained by collecting more days alone and is consistent with allocating the days to more valuable homes and periods.

cs.LG

版本更新 · v3arXiv:2503.22821

Identifying and Mitigating API Misuse in Large Language Models

Terry Yue Zhuo, Junda He, Jiamou Sun, Zhenchang Xing, David Lo, John Grundy, Xiaoning Du

原文摘要

API misuse in code generated by large language models (LLMs) presents a serious and growing challenge in software development, as although LLMs demonstrate impressive code generation capabilities, their interactions with complex library APIs are often error-prone and can lead to software failures and vulnerabilities. In this paper, we conduct a large-scale study of API misuse patterns in LLM-generated code by analyzing both method selection and parameter usage across Python and Java, using three representative LLMs: StarCoder-7B, Qwen2.5-Coder-7B, and GitHub Copilot. Based on extensive manual annotation of 3,209 method-level and 3,492 parameter-level misuses, we identify and categorize four recurring misuse types by building on and refining prior API misuse taxonomies. Our evaluation of the three LLMs reveals persistent challenges in API usage, particularly hallucination and intent misalignment. To address these issues, we propose Dr.Fix, an LLM-based automatic repair approach guided by our taxonomy, which improves repair accuracy compared to baseline prompting and existing repair methods, achieving gains of up to 38.4 BLEU and up to 72 more exact matches on benchmark datasets. This work offers important insights into the current limitations of LLMs in API usage and points to directions for improving automated misuse repair in code generation systems.

cs.SE

新论文 · v1arXiv:2611.17950

Toward Interference-Agnostic RF Source Separation

Maxence Albert, Raphaël Duroselle, Thibaud Ehret, Olivier Boeffard, Jean-François Bonastre

原文摘要

Separating a radio-frequency signal of interest from a co-channel interference is mostly addressed with specialized separators, each trained for one pair of signal and interference types and therefore relying on a label that a heavily jammed receiver cannot identify. We show that this reliance is not benign: applied to an interference it was not trained on, such a separator usually fails to recover the signal, and in most of the cases we measure it leaves the signal worse than no processing at all. We then train a single separator over every pair, giving it no indication at test time of which one it faces, on two architectures with no component in common: the convolutional WaveNet of the ICASSP 2024 RF Challenge, and a band-split transformer that we adapt from music source separation. Both behave the same way. At the size of one specialized model, the unified separator stays close to the whole models it replaces, keeping the bit error rate close on almost every pair, better on one and worse on another. On an interference absent from every training set, it does as well as the best specialized model, without anyone having to choose which one to deploy.

cs.AI · eess.SP

新论文 · v1arXiv:2611.17939

Can You Trust Your Finance Agent? Drift, Diagnosis and Auditable Intervention on One Shipped Financial Research Agent

Cosimo Galeone, Anna Ettorre, Ginevra Mariani, Minsu Park, Giuseppe Ettorre, Daniele Ligorio

原文摘要

Can you trust a finance agent you did not build? We take the reference agent shipped with the Vals Finance Agent Benchmark v2, leave it untuned, and ask four questions a leaderboard score does not answer. Is the score stable? Seven identical runs on 18 questions pass between 105 and 123 of 149 rubric criteria, and the bill varies by up to 17%. When the agent fails, who owns the fix? We classify 20 failed trajectories, from three runs of nine held-out questions, by the layer whose owner would remedy them, under a taxonomy frozen in advance: 13 belong to the agent's instructions, 3 to its tools, 4 to the model, none to its data. Can that diagnosis be trusted? Four blind readers from three model families, given only the traces, name the same fault as our LLM analyzer in 94% of cases, the rate at which they agree with one another (93%). Can a change be trusted? An optimization loop reads the traces, makes one targeted change per round and keeps it only if it clears a fixed threshold. Under a cost objective its final configuration costs 0.34x the baseline on the held-out questions (95% CI [0.22, 0.58]) at accuracy within noise. Under an accuracy objective it gains 14 points on the development questions and 1.5 on the held-out ones (CI [-5.3, +7.2]). Judged against the measured run-to-run noise, under a rule written before the test was opened, the first is a real saving and the second is not. This is a case study of one agent and nine test questions, not a general result.

cs.CL · cs.AI

版本更新 · v4arXiv:2610.23170

Chinese Decision: A Family of Language Models for Single-Pass Probabilistic Decision-Making

Qiang Yan, Jiafeng Guo

原文摘要

Many applications need either a probability distribution over a few decisions or a set of typed spans from a text, and both are usually obtained by generating text. We present Chinese Decision, a family of Chinese models that answers both with a single forward pass and no generated token: candidate-letter logits give probabilities for Boolean, categorical, and ordinal decisions, and a small span head reading an echoed copy of the input gives typed character spans. Both readouts share one backbone and one merged LoRA, so a unified model is one weight file. We release decision-only and unified models at 0.8B, 2B, and 4B (Qwen3.5), trained for one epoch on 13.05 million decisions and, for the unified models, on CJ-IE, a 1.2-million-row Chinese extraction mixture. On the full 307,900-decision CJ-Bench, the unified 4B model reaches 86.86% accuracy, above every reported baseline and the decision-only 4B model (85.97%), and extracts spans on CJ-IE with micro F1 of 68.9. Retention has a price: against a 0.8B model continued on CJ-IE alone, which reaches 73.9 F1 but forgets CJ-Bench (71.02%), keeping decision ability costs 5.0 to 10.7 F1 points, shrinking with size. Gains of the unified models over decision-only ones are small, and all comparisons rest on single training runs with paired clustered bootstrap intervals.

cs.CL

版本更新 · v2arXiv:2610.27508

KatoANN: Enabling Near-Exact Vector Search at the Speed of Lossy Scans with CPU Matrix Engines

Soobin Cho, Gyuyeong Kim

原文摘要

Emerging CPU matrix engines, such as Intel AMX, enable high throughput by batching multiple queries. However, they support only low-precision formats, which can significantly degrade recall (i.e., search quality). We present KatoANN, the first approximate nearest neighbor (ANN) search system that achieves both high throughput and high recall simultaneously on CPU matrix engines. Specifically, KatoANN represents each vector using a 16-bit code, splits it into high- and low-byte planes, and computes near-exact distances using 8-bit matrix multiplications that the matrix engine supports. To reduce extra multiplication overhead, it ranks all candidates by their high bytes alone and refines only a few survivors with the full distance. KatoANN also reduces memory footprint since it does not maintain a full-precision copy of the vectors. We implement a KatoANN prototype on Intel AMX. Our experimental results show that KatoANN outperforms existing partition-based systems across the high-recall range.

cs.DB · cs.AR

版本更新 · v2arXiv:2610.20579

PingpongDuo: Agile Humanoid Table Tennis Doubles with Safe Team Coordination

Haoran Jiang, Yichao Zhong, Kai Zhang, Cheng Liu, Yixuan Pan, Zhenjie Yang, Jingbo Wang, Tianyu Li 等 9 位作者

原文摘要

Unlike singles, humanoid table tennis doubles requires teammates to strike in strict alternation while sharing a confined workspace. Rapid repositioning can compromise striking accuracy, while uncoordinated teammate movements introduce collision risks. We present PingpongDuo, a hierarchical framework for agile humanoid doubles with safe team coordination. At the team level, a planner jointly selects positioning targets and uses learned risk estimates to filter risky proposals, preferentially preserving the current striker's target. At the individual level, separate footwork and striking skills combine agile movement with precise racket control. The skills are guided by human motion references, including footwork collected from human doubles play, and trained from states produced by one another to support rapid transitions. A racket-arm residual further refines strikes after movement using return feedback from a simulator calibrated with real ball-flight and contact measurements. In simulation, PingpongDuo improves the ball-return success rate by 13.2% over the state-of-the-art baseline, while achieving approximately 2 cm positioning error during clearance. In real-world experiments, a robot completes a 2.15 m clearance-and-recovery cycle in 2.4 s with a peak speed of 2.23 m/s, while two robots sustain more than 10 consecutive strikes in strict alternation.

cs.RO

版本更新 · v2arXiv:2611.16408

Gravity and Flight: Learning to Navigate Fine-Grained Remote Sensing Scenes

Yuheng Zong, Minghua Wang, Xin Zhao, Xiaobin Liu, Yongjian Deng, Chanho Eom

原文摘要

Vision-language models are increasingly capable, yet their reasoning in complex remote sensing scenes can remain coarse-grained and insufficiently grounded in task-relevant visual evidence. We propose Gravity and Flight (G&F), a framework for scene specialization with limited supervision that separates visual and reasoning verification while coupling policy updates. It combines Evidence-of-Thought supervised fine-tuning (EoT-SFT) and Relational Transfer Policy Optimization (RTPO). EoT-SFT teaches fine-grained observation and faithful reasoning through description, analysis, and supported conclusions. RTPO independently samples responses to scene-matched image pairs with fine-grained visual differences under a shared task. Its rubric-based verification combines Gravity, which checks visual fidelity and description coverage, and Flight, which takes descriptions as premises to assess evidence use and relations between paired judgments. Their scores and task correctness yield a shared pair-level advantage that couples policy updates from both responses. Using 1,008 demonstrations and 500 training pairs for harbor specialization, G&F achieves a mean score of 69.12 across six evaluation sets, exceeding the backbone and supervised initialization by 9.76 and 4.56 points. Matched controls show gains over single-image and joint-input RL, and removing either verification view reduces performance. Applying G&F to urban specialization also improves on the base model and supervised initialization across three evaluation subsets. Project details and updates are available at https://github.com/Z0ngL1ng/Gravity-and-Flight.

cs.CV

版本更新 · v2arXiv:2611.14937

Caprese: Change Impact Recommendation Using History and Runtime Analysis

Sadjad Tavakoli, Saba Alimadadi

原文摘要

Change impact recommendation identifies program entities potentially affected by source code changes. Historical co-changes and runtime dependencies provide complementary evidence, particularly in JavaScript, where dynamic and asynchronous behaviour complicates dependency analysis. We present Caprese, an open-source framework for function-level change impact recommendation in Node.js. Given a change set, it produces a ranked impact set of potentially affected functions. It supports repository-based recommendation through confidence-guided mining of recurring, change-set-relevant co-change patterns of varying lengths; dynamic recommendation through execution-derived dependencies involving calls, callbacks, events, timers, and shared state; and hybrid recommendation integrating both analyses. Its modular implementation supports reusable analyses. Evaluation on ten open-source applications and 100 selected change sets shows that, on average across projects, 77.4% of the relevant functions among the top 60 hybrid recommendations are identified exclusively by one analysis, demonstrating substantial complementarity. Dynamic recommendation achieves the highest mean precision among the top five recommendations (0.65), while repository-based recommendation performs best when the ranking is extended to 60 recommendations (0.42). Hybrid recommendation outperforms TARMAQ on average but does not consistently outperform the individual modes. Caprese is publicly available with source code, documentation, and evaluation artifacts.

cs.SE

新论文 · v1arXiv:2611.17334

拆解多模态视觉编码

Unbundling Multimodal Vision Encoding

Ellis Brown, Chris Hoang, Shengyi Qian, Shang-Wen Li, Mike Lewis, Marjan Ghazvininejad, Luke Zettlemoyer, Lucas Beyer 等 12 位作者

AI 速读 基于摘要

在统一数据与匹配总计算量的受控框架下,作者报告视觉先验并非必需,真正关键的是专用视觉容量,且该容量可置于多模态融合早期而非融合前。

cs.CV · cs.LG

查看原文摘要

How multimodal language models should encode visual input remains up for debate. Most models use a pretrained vision encoder, while some recent recipes train the encoder from scratch with the language model or remove it entirely. Vision encoders bundle together three design choices: a visual prior from upstream pretraining, dedicated vision capacity, and pre-fusion placement of that capacity (i.e., before the multimodal backbone). Because published models also differ in data, architecture, and compute, comparing them cannot isolate the effect of each choice. We disentangle these choices via a controlled framework where we pretrain all models from scratch on common data and compare them on visual question answering and language benchmarks at matched lifetime compute. We find that the visual prior is not necessary; an encoder trained jointly from scratch is on par with a pretrained one, without a separate training stage. What matters instead is dedicated vision capacity: an encoder-free model trails on visual understanding even when its backbone is enlarged by the encoder's parameter count. Yet that capacity need not precede multimodal fusion, as vision-only modules added in-fusion to early backbone layers perform comparably to a pre-fusion encoder and can lower time to first token. These results also hold across backbone sizes, when encoder pretraining is treated as free, and when starting from a pretrained language model. Overall, our findings favor dedicated vision capacity trained jointly with the backbone and placed early within it.