cs.AI（2026-03-17）

📊 共 43 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (32 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (8 🔗1) 支柱一：机器人控制 (Robot Control) (2) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (32 篇)

#题目一句话要点标签🔗⭐
1 Enhancing Linguistic Generalization of VLA: Fine-Tuning OpenVLA via Synthetic Instruction Augmentation 通过合成指令增强微调OpenVLA，提升具身AI的语言泛化能力 embodied AI vision-language-action VLA
2 Surg$Σ$: A Spectrum of Large-Scale Multimodal Data and Foundation Models for Surgical Intelligence Surg$Σ$: 构建大规模多模态手术数据集与模型，提升手术智能跨任务泛化能力。 large language model foundation model multimodal
3 ExpressMind: A Multimodal Pretrained Large Language Model for Expressway Operation 提出ExpressMind以解决高速公路智能运营问题 large language model multimodal chain-of-thought ✅
4 InCoder-32B: Code Foundation Model for Industrial Scenarios InCoder-32B：面向工业场景的代码大模型，统一多领域代码智能 large language model foundation model
5 Structure-Aware Multimodal LLM Framework for Trustworthy Near-Field Beam Prediction 提出结构感知多模态LLM框架，用于可信的近场波束预测 large language model multimodal
6 Prompt Programming for Cultural Bias and Alignment of Large Language Models 提出基于DSPy的提示编程方法，用于优化大语言模型的文化偏见与对齐 large language model
7 From Natural Language to Executable Option Strategies via Large Language Models 提出基于大语言模型的神经符号方法，将自然语言转化为可执行期权策略 large language model
8 Detecting Sentiment Steering Attacks on RAG-enabled Large Language Models 提出基于CNN和LSTM的轻量级入侵检测系统，增强物联网网络安全。 large language model
9 A Human-Centred Architecture for Large Language Models-Cognitive Assistants in Manufacturing within Quality Management Systems 提出一种以人为本的LLM-CA架构，用于增强制造质量管理系统 large language model
10 Are Large Language Models Truly Smarter Than Humans? 多方法污染审计揭示大型语言模型在公开基准测试中存在数据污染问题 large language model
11 Resource Consumption Threats in Large Language Models 综述性研究：系统性分析大语言模型中的资源消耗威胁及其应对 large language model
12 NeSy-Route: A Neuro-Symbolic Benchmark for Constrained Route Planning in Remote Sensing NeSy-Route：遥感约束路径规划的神经符号基准测试 large language model multimodal
13 Diffusion Models for Joint Audio-Video Generation 提出基于扩散模型的联合音视频生成方法，并构建高质量数据集。 multimodal
14 Is Conformal Factuality for RAG-based LLMs Robust? Novel Metrics and Systematic Insights 分析RAG中Conformal Factuality的鲁棒性，提出新指标并揭示其局限性 large language model
15 Learning to Predict, Discover, and Reason in High-Dimensional Discrete Event Sequences 提出基于Transformer的框架，用于预测、发现和推理高维离散事件序列，解决汽车故障诊断难题。 large language model
16 SocialOmni: Benchmarking Audio-Visual Social Interactivity in Omni Models SocialOmni：提出用于评估Omni模型在音视频社交互动能力的基准 large language model
17 Internalizing Agency from Reflective Experience LEAFE：通过反思经验内化行动能力，提升LLM智能体长程任务问题解决能力 large language model
18 Differential Harm Propensity in Personalized LLM Agents: The Curious Case of Mental Health Disclosure 研究用户心理健康披露对个性化LLM Agent有害行为的影响，揭示安全-效用权衡。 large language model
19 IQuest-Coder-V1 Technical Report IQuest-Coder-V1：提出代码流多阶段训练范式，提升代码大语言模型在软件工程、编程竞赛和工具使用上的性能。 large language model
20 When AI Navigates the Fog of War 利用LLM在“战争迷雾”中进行地缘政治预测：一项前瞻性分析 large language model
21 Runtime Governance for AI Agents: Policies on Paths 提出AI Agent运行时治理框架，通过路径策略实现动态合规控制 large language model
22 BenchPreS: A Benchmark for Context-Aware Personalized Preference Selectivity of Persistent-Memory LLMs BenchPreS：评估持久内存LLM在上下文感知下的个性化偏好选择性 large language model
23 Exploring different approaches to customize language models for domain-specific text-to-code generation 探索定制化语言模型用于领域特定文本到代码生成的不同方法 large language model
24 RetailBench: Evaluating Long-Horizon Autonomous Decision-Making and Strategy Stability of LLM Agents in Realistic Retail Environments RetailBench：评估LLM智能体在零售环境中长期自主决策与策略稳定性 large language model
25 An Efficient Heterogeneous Co-Design for Fine-Tuning on a Single GPU SlideFormer：一种高效异构协同设计，用于在单GPU上微调大型语言模型 large language model
26 Toward Experimentation-as-a-Service in 5G/6G: The Plaza6G Prototype for AI-Assisted Trials Plaza6G：面向5G/6G的实验即服务平台，支持AI辅助试验 large language model
27 Adaptive Theory of Mind for LLM-based Multi-Agent Coordination 提出自适应心智理论以解决多智能体协调问题 large language model
28 CoMAI: A Collaborative Multi-Agent Framework for Robust and Equitable Interview Evaluation CoMAI：用于稳健和公平面试评估的协同多智能体框架 large language model
29 MOSAIC: Composable Safety Alignment with Modular Control Tokens MOSAIC：通过模块化控制令牌实现可组合的安全对齐 large language model
30 Proactive Rejection and Grounded Execution: A Dual-Stage Intent Analysis Paradigm for Safe and Efficient AIoT Smart Homes 提出双阶段意图感知框架，提升AIoT智能家居的安全性和效率 large language model
31 Efficient LLM Serving for Agentic Workflows: A Data Systems Perspective Helium：面向Agent工作流的高效LLM服务框架，优化跨调用依赖 large language model
32 A Context Alignment Pre-processor for Enhancing the Coherence of Human-LLM Dialog 提出上下文对齐预处理器C.A.P.，增强人机对话中LLM的连贯性。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
33 Follow the Clues, Frame the Truth: Hybrid-evidential Deductive Reasoning in Open-Vocabulary Multimodal Emotion Recognition 提出HyDRA，通过混合证据演绎推理解决开放词汇多模态情感识别中的歧义性问题 reinforcement learning reward shaping open-vocabulary
34 Anticipatory Planning for Multimodal AI Agents TraceR1：通过预测轨迹进行前瞻性规划，提升多模态AI Agent的决策能力 reinforcement learning multimodal
35 Via Negativa for AI Alignment: Why Negative Constraints Are Structurally Superior to Positive Preferences 提出基于负约束的AI对齐方法，解决偏好学习中的谄媚问题。 reinforcement learning PPO RLHF
36 Multi-Agent Reinforcement Learning Counteracts Delayed CSI in Multi-Satellite Systems 提出DS-PPO算法，解决多卫星系统中因信道状态信息延迟导致的速率优化问题 reinforcement learning PPO
37 ARISE: Agent Reasoning with Intrinsic Skill Evolution in Hierarchical Reinforcement Learning 提出ARISE框架，通过内在技能演化提升Agent在数学推理中的能力 reinforcement learning reward design ✅
38 What if Pinocchio Were a Reinforcement Learning Agent: A Normative End-to-End Pipeline 提出Pino：一个基于论证的规范强化学习端到端流程，解决智能体规范遵从问题 reinforcement learning
39 TRUST-SQL: Tool-Integrated Multi-Turn Reinforcement Learning for Text-to-SQL over Unknown Schemas 提出TRUST-SQL，解决未知Schema下的Text-to-SQL问题，无需预加载元数据。 reinforcement learning
40 SWE-QA-Pro: A Representative Benchmark and Scalable Training Recipe for Repository-Level Code Understanding SWE-QA-Pro：提出代码仓库级理解的代表性基准和可扩展训练方案。 reinforcement learning large language model

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
41 Visual Distraction Undermines Moral Reasoning in Vision-Language Models 视觉干扰削弱视觉-语言模型中的道德推理能力 manipulation multimodal
42 CritiSense: Critical Digital Literacy and Resilience Against Misinformation CritiSense：多语言数字素养App，提升用户抵御虚假信息能力 manipulation

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
43 LenghuSky-8: An 8-Year All-Sky Cloud Dataset with Star-Aware Masks and Alt-Az Calibration for Segmentation and Nowcasting LenghuSky-8：用于云分割和临近预报的八年全天云数据集，包含星敏感掩膜和Alt-Az校准 optical flow

⬅️ 返回 cs.AI 首页 · 🏠 返回主页