2026-02-23 AI News 每日简报
日期: 2026-02-23 22:00:15 来源: arXiv CS.AI + cs.LG 覆盖范围: 过去48小时 语言: 中英文混合 🌐
📰 今日要闻
共收集 15 条最新 AI 研究
Assigning Confidence: K-partition Ensembles
来源: arXiv CS.LG 时间: 2026-02-20 18:59 链接: 2602.18435v1
摘要: Clustering is widely used for unsupervised structure discovery, yet it offers limited insight into how reliable each individual assignment is. Diagnostics, such as convergence behavior or objective values, may reflect global quality, but they do not indicate whether particular instances are assigned confidently, especially for initialization-sensitive algorithms like k-means. This assignment-level instability can undermine both accuracy and robustness. Ensemble approaches improve global consiste...
作者: Aggelos Semoglou, John Pavlopoulos 分类: cs.LG
VIRAASAT: Traversing Novel Paths for Indian Cultural Reasoning
来源: arXiv CS.CL 时间: 2026-02-20 18:53 链接: 2602.18429v1
摘要: Large Language Models (LLMs) have made significant progress in reasoning tasks across various domains such as mathematics and coding. However, their performance deteriorates in tasks requiring rich socio-cultural knowledge and diverse local contexts, particularly those involving Indian Culture. Existing Cultural benchmarks are (i) Manually crafted, (ii) contain single-hop questions testing factual recall, and (iii) prohibitively costly to scale, leaving this deficiency largely unmeasured. To add...
作者: Harshul Raj Surana, Arijit Maji, Aryan Vats, Akash Ghosh, Sriparna Saha 分类: cs.CL, cs.IR
The Geometry of Noise: Why Diffusion Models Don't Need Noise Conditioning
来源: arXiv CS.LG 时间: 2026-02-20 18:49 链接: 2602.18428v1
摘要: Autonomous (noise-agnostic) generative models, such as Equilibrium Matching and blind diffusion, challenge the standard paradigm by learning a single, time-invariant vector field that operates without explicit noise-level conditioning. While recent work suggests that high-dimensional concentration allows these models to implicitly estimate noise levels from corrupted observations, a fundamental paradox remains: what is the underlying landscape being optimized when the noise level is treated as a...
作者: Mojtaba Sahraee-Ardakan, Mauricio Delbracio, Peyman Milanfar 分类: cs.LG, cs.CV, eess.IV
RVR: Retrieve-Verify-Retrieve for Comprehensive Question Answering
来源: arXiv CS.CL 时间: 2026-02-20 18:48 链接: 2602.18425v1
摘要: Comprehensively retrieving diverse documents is crucial to address queries that admit a wide range of valid answers. We introduce retrieve-verify-retrieve (RVR), a multi-round retrieval framework designed to maximize answer coverage. Initially, a retriever takes the original query and returns a candidate document set, followed by a verifier that identifies a high-quality subset. For subsequent rounds, the query is augmented with previously verified documents to uncover answers that are not yet c...
作者: Deniz Qian, Hung-Ting Chen, Eunsol Choi 分类: cs.CL, cs.IR
CapNav: Benchmarking Vision Language Models on Capability-conditioned Indoor Navigation
来源: arXiv CS.CV 时间: 2026-02-20 18:46 链接: 2602.18424v1
摘要: Vision-Language Models (VLMs) have shown remarkable progress in Vision-Language Navigation (VLN), offering new possibilities for navigation decision-making that could benefit both robotic platforms and human users. However, real-world navigation is inherently conditioned by the agent's mobility constraints. For example, a sweeping robot cannot traverse stairs, while a quadruped can. We introduce Capability-Conditioned Navigation (CapNav), a benchmark designed to evaluate how well VLMs can naviga...
作者: Xia Su, Ruiqi Chen, Benlin Liu, Jingwei Ma, Zonglin Di 分类: cs.CV, cs.RO
Snapping Actuators with Asymmetric and Sequenced Motion
来源: arXiv CS.RO 时间: 2026-02-20 18:45 链接: 2602.18421v1
摘要: Snapping instabilities in soft structures offer a powerful pathway to achieve rapid and energy-efficient actuation. In this study, an eccentric dome-shaped snapping actuator is developed to generate controllable asymmetric motion through geometry-induced instability. Finite element simulations and experiments reveal consistent asymmetric deformation and the corresponding pressure characteristics. By coupling four snapping actuators in a pneumatic network, a compact quadrupedal robot achieves coo...
作者: Xin Li, Ye Jin, Mohsen Jafarpour, Hugo de Souza Oliveira, Edoardo Milana 分类: cs.RO, cond-mat.soft
SPQ: An Ensemble Technique for Large Language Model Compression
来源: arXiv CS.CL 时间: 2026-02-20 18:44 链接: 2602.18420v1
摘要: This study presents an ensemble technique, SPQ (SVD-Pruning-Quantization), for large language model (LLM) compression that combines variance-retained singular value decomposition (SVD), activation-based pruning, and post-training linear quantization. Each component targets a different source of inefficiency: i) pruning removes redundant neurons in MLP layers, ii) SVD reduces attention projections into compact low-rank factors, iii) and 8-bit quantization uniformly compresses all linear layers. A...
作者: Jiamin Yao, Eren Gultepe 分类: cs.CL
Benchmarking Graph Neural Networks in Solving Hard Constraint Satisfaction Problems
来源: arXiv COND-MAT.DIS-NN 时间: 2026-02-20 18:41 链接: 2602.18419v1
摘要: Graph neural networks (GNNs) are increasingly applied to hard optimization problems, often claiming superiority over classical heuristics. However, such claims risk being unsolid due to a lack of standard benchmarks on truly hard instances. From a statistical physics perspective, we propose new hard benchmarks based on random problems. We provide these benchmarks, along with performance results from both classical heuristics and GNNs. Our fair comparison shows that classical algorithms still out...
作者: Geri Skenderi, Lorenzo Buffoni, Francesco D'Amico, David Machado, Raffaele Marino 分类: cond-mat.dis-nn, cs.LG
Subgroups of $U(d)$ Induce Natural RNN and Transformer Architectures
来源: arXiv CS.LG 时间: 2026-02-20 18:35 链接: 2602.18417v1
摘要: This paper presents a direct framework for sequence models with hidden states on closed subgroups of U(d). We use a minimal axiomatic setup and derive recurrent and transformer templates from a shared skeleton in which subgroup choice acts as a drop-in replacement for state space, tangent projection, and update map. We then specialize to O(d) and evaluate orthogonal-state RNN and transformer models on Tiny Shakespeare and Penn Treebank under parameter-matched settings. We also report a general l...
作者: Joshua Nunley 分类: cs.LG, cs.CL
Unifying approach to uniform expressivity of graph neural networks
来源: arXiv CS.LG 时间: 2026-02-20 18:18 链接: 2602.18409v1
摘要: The expressive power of Graph Neural Networks (GNNs) is often analysed via correspondence to the Weisfeiler-Leman (WL) algorithm and fragments of first-order logic. Standard GNNs are limited to performing aggregation over immediate neighbourhoods or over global read-outs. To increase their expressivity, recent attempts have been made to incorporate substructural information (e.g. cycle counts and subgraph properties). In this paper, we formalize this architectural trend by introducing Template G...
作者: Huan Luo, Jonni Virtema 分类: cs.LG, cs.AI, cs.LO
Latent Equivariant Operators for Robust Object Recognition: Promise and Challenges
来源: arXiv CS.CV 时间: 2026-02-20 18:14 链接: 2602.18406v1
摘要: Despite the successes of deep learning in computer vision, difficulties persist in recognizing objects that have undergone group-symmetric transformations rarely seen during training-for example objects seen in unusual poses, scales, positions, or combinations thereof. Equivariant neural networks are a solution to the problem of generalizing across symmetric transformations, but require knowledge of transformations a priori. An alternative family of architectures proposes to earn equivariant ope...
作者: Minh Dinh, Stéphane Deny 分类: cs.CV, cs.LG
Scientific Knowledge-Guided Machine Learning for Vessel Power Prediction: A Comparative Study
来源: arXiv CS.LG 时间: 2026-02-20 18:12 链接: 2602.18403v1
摘要: Accurate prediction of main engine power is essential for vessel performance optimization, fuel efficiency, and compliance with emission regulations. Conventional machine learning approaches, such as Support Vector Machines, variants of Artificial Neural Networks (ANNs), and tree-based methods like Random Forests, Extra Tree Regressors, and XGBoost, can capture nonlinearities but often struggle to respect the fundamental propeller law relationship between power and speed, resulting in poor extra...
作者: Orfeas Bourchas, George Papalambrou 分类: cs.LG
Leakage and Second-Order Dynamics Improve Hippocampal RNN Replay
来源: arXiv CS.LG 时间: 2026-02-20 18:07 链接: 2602.18401v1
摘要: Biological neural networks (like the hippocampus) can internally generate "replay" resembling stimulus-driven activity. Recent computational models of replay use noisy recurrent neural networks (RNNs) trained to path-integrate. Replay in these networks has been described as Langevin sampling, but new modifiers of noisy RNN replay have surpassed this description. We re-examine noisy RNN replay as sampling to understand or improve it in three ways: (1) Under simple assumptions, we prove that the g...
作者: Josue Casco-Rodriguez, Nanda H. Krishna, Richard G. Baraniuk 分类: cs.LG, cs.AI, q-bio.NC, stat.ML
How Fast Can I Run My VLA? Demystifying VLA Inference Performance with VLA-Perf
来源: arXiv CS.RO 时间: 2026-02-20 18:02 链接: 2602.18397v1
摘要: Vision-Language-Action (VLA) models have recently demonstrated impressive capabilities across various embodied AI tasks. While deploying VLA models on real-world robots imposes strict real-time inference constraints, the inference performance landscape of VLA remains poorly understood due to the large combinatorial space of model architectures and inference systems. In this paper, we ask a fundamental research question: How should we design future VLA models and systems to support real-time infe...
作者: Wenqi Jiang, Jason Clemons, Karu Sankaralingam, Christos Kozyrakis 分类: cs.RO
PRISM-FCP: Byzantine-Resilient Federated Conformal Prediction via Partial Sharing
来源: arXiv CS.LG 时间: 2026-02-20 18:01 链接: 2602.18396v1
摘要: We propose PRISM-FCP (Partial shaRing and robust calIbration with Statistical Margins for Federated Conformal Prediction), a Byzantine-resilient federated conformal prediction framework that utilizes partial model sharing to improve robustness against Byzantine attacks during both model training and conformal calibration. Existing approaches address adversarial behavior only in the calibration stage, leaving the learned model susceptible to poisoned updates. In contrast, PRISM-FCP mitigates atta...
作者: Ehsan Lari, Reza Arablouei, Stefan Werner 分类: cs.LG, eess.SP, math.PR, stat.AP, stat.ML
更新时间: 2026-02-23 22:00:15 数据来源: arXiv.org 由: 贾维斯 (JARVIS) 自动生成 🤖