cs.AI（2025-12-23）

📊 共 29 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) (18 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (9 🔗1) 支柱一：机器人控制 (Robot Control) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (18 篇)

#题目一句话要点标签🔗⭐
1 Odysseus: Jailbreaking Commercial Multimodal LLM-integrated Systems via Dual Steganography Odysseus：利用双重隐写术破解商业多模态LLM集成系统 large language model multimodal
2 Automated stereotactic radiosurgery planning using a human-in-the-loop reasoning large language model agent SAGE：基于人机协同推理的大语言模型自动立体定向放射外科计划系统 large language model chain-of-thought
3 Dual-Encoder Transformer-Based Multimodal Learning for Ischemic Stroke Lesion Segmentation Using Diffusion MRI 提出基于双编码器Transformer的Ischemic Stroke病灶分割方法，提升DWI和ADC图像的分割精度。 multimodal
4 Toward Explaining Large Language Models in Software Engineering Tasks 提出FeatureSHAP，用于解释软件工程任务中的大型语言模型 large language model ✅
5 Advancing Multimodal Teacher Sentiment Analysis:The Large-Scale T-MED Dataset & The Effective AAM-TSA Model 构建T-MED数据集与AAM-TSA模型以提升教师情感分析准确性 multimodal
6 SynCraft: Guiding Large Language Models to Predict Edit Sequences for Molecular Synthesizability Optimization SynCraft：引导大语言模型预测编辑序列，优化分子合成可行性 large language model
7 TongSIM: A General Platform for Simulating Intelligent Machines TongSIM：通用智能机器模拟平台，支持具身智能体训练与评估 embodied AI large language model multimodal
8 Concept Generalization in Humans and Large Language Models: Insights from the Number Game 通过数字游戏对比人类与大语言模型在概念泛化上的差异 large language model
9 A DeepSeek-Powered AI System for Automated Chest Radiograph Interpretation in Clinical Practice DeepSeek赋能的AI系统Janus-Pro-CXR，用于临床胸部X光片自动判读 large language model multimodal
10 Reason2Decide: Rationale-Driven Multi-Task Learning Reason2Decide：一种基于理由驱动的多任务学习框架，提升临床决策支持系统的预测精度和解释一致性。 large language model foundation model
11 Generative Digital Twins: Vision-Language Simulation Models for Executable Industrial Systems 提出视觉-语言模拟模型，从草图和文本生成可执行的工业系统数字孪生。 multimodal
12 Synthesizing Procedural Memory: Challenges and Architectures in Automated Workflow Generation 提出一种自动工作流生成方法，解决大型语言模型从工具使用者到工作流架构师的转变难题。 large language model
13 Memory as Resonance: A Biomimetic Architecture for Infinite Context Memory on Ergodic Phonetic Manifolds 提出基于遍历语音流形的共振记忆架构PTM，解决大语言模型无限上下文记忆问题。 large language model
14 MemR$^3$: Memory Retrieval via Reflective Reasoning for LLM Agents MemR³：通过反思推理实现LLM Agent的记忆检索，提升问答质量。 large language model
15 AXIOM: Benchmarking LLM-as-a-Judge for Code via Rule-Based Perturbation and Multisource Quality Calibration AXIOM：通过规则扰动和多源质量校准，基准测试LLM作为代码评估判官的能力 large language model
16 Enhancing Zero-Shot Time Series Forecasting in Off-the-Shelf LLMs via Noise Injection 通过噪声注入增强即用型LLM的零样本时间序列预测能力 large language model
17 On the Effectiveness of Instruction-Tuning Local LLMs for Identifying Software Vulnerabilities 指令调优本地LLM，有效识别软件漏洞类型，提升安全性和实用性。 large language model
18 S$^3$IT: A Benchmark for Spatially Situated Social Intelligence Test 提出S$^3$IT基准测试，用于评估具身智能体在复杂社交环境中的推理能力 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (9 篇)

#题目一句话要点标签🔗⭐
19 Scaling Reinforcement Learning for Content Moderation with Large Language Models 利用强化学习和大型语言模型提升大规模内容审核的效率与准确性 reinforcement learning reward shaping large language model
20 Identifying Appropriately-Sized Services with Deep Reinforcement Learning 提出Rake，利用深度强化学习从实现工件中识别合适大小的服务。 reinforcement learning deep reinforcement learning
21 Adaptive Financial Sentiment Analysis for NIFTY 50 via Instruction-Tuned LLMs , RAG and Reinforcement Learning Approaches 提出基于指令调优LLM、RAG和强化学习的自适应金融情感分析框架，用于NIFTY 50指数预测。 reinforcement learning PPO large language model
22 Graph-Symbolic Policy Enforcement and Control (G-SPEC): A Neuro-Symbolic Framework for Safe Agentic AI in 5G Autonomous Networks 提出G-SPEC神经符号框架，保障5G自治网络中LLM代理的安全策略执行。 reinforcement learning deep reinforcement learning large language model
23 LongVideoAgent: Multi-Agent Reasoning with Long Videos LongVideoAgent：提出一种基于多智能体推理的长视频问答框架，提升时序定位和细节捕捉能力。 reinforcement learning multimodal ✅
24 Leveraging High-Fidelity Digital Models and Reinforcement Learning for Mission Engineering: A Case Study of Aerial Firefighting Under Perfect Information 利用高保真数字模型与强化学习进行任务工程：以完美信息下的空中消防为例 reinforcement learning
25 Offline Safe Policy Optimization From Heterogeneous Feedback 提出PreSa框架，通过异构反馈直接优化安全策略，解决离线安全策略优化问题 reinforcement learning preference learning RLHF
26 Evolutionary Neural Architecture Search with Dual Contrastive Learning 提出DCL-ENAS，利用双重对比学习提升进化神经架构搜索的效率和精度。 contrastive learning
27 Discovering Lie Groups with Flow Matching 提出流匹配方法以发现李群的对称性 flow matching

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
28 ActionFlow: A Pipelined Action Acceleration for Vision Language Models on Edge ActionFlow：边缘设备上视觉语言模型流水线式动作加速框架 manipulation vision-language-action VLA

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

#题目一句话要点标签🔗⭐
29 Learning Skills from Action-Free Videos 提出基于光流的技能抽象框架SOF，从无动作视频中学习机器人技能 optical flow

⬅️ 返回 cs.AI 首页 · 🏠 返回主页