cs.LG（2026-01-19）

📊 共 20 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) (9) 支柱二：RL算法与架构 (RL & Architecture) (9 🔗2) 支柱五：交互与反应 (Interaction & Reaction) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (9 篇)

#题目一句话要点标签🔗⭐
1 FastAV: Efficient Token Pruning for Audio-Visual Large Language Model Inference FastAV：面向音视频大语言模型推理的高效Token剪枝框架 large language model multimodal
2 Semi-supervised Instruction Tuning for Large Language Models on Text-Attributed Graphs 提出半监督指令调优方法以解决图学习中的标签稀缺问题 large language model
3 A Comprehensive Evaluation of LLM Reasoning: From Single-Model to Multi-Agent Paradigms 全面评估LLM推理范式：从单模型到多智能体系统，揭示其性能与成本权衡。 large language model chain-of-thought
4 PASs-MoE: Mitigating Misaligned Co-drift among Router and Experts via Pathway Activation Subspaces for Continual Learning 提出PASs-MoE，通过路径激活子空间缓解持续学习中路由与专家之间的错位共漂移问题 large language model multimodal
5 The Tag is the Signal: URL-Agnostic Credibility Scoring for Messages on Telegram 提出TAG2CRED模型，通过标签分析提升Telegram消息可信度评估，尤其针对短文本和URL稀疏消息。 large language model
6 Polychronous Wave Computing: Timing-Native Address Selection in Spiking Networks 提出多时波计算，实现脉冲神经网络中基于时序的原生地址选择。 TAMP
7 CooperLLM: Cloud-Edge-End Cooperative Federated Fine-tuning for LLMs via ZOO-based Gradient Correction CooperLLM：基于ZOO梯度校正的云边端协同联邦微调LLM large language model
8 PDFInspect: A Unified Feature Extraction Framework for Malicious Document Detection PDFInspect：用于恶意文档检测的统一特征提取框架 TAMP
9 MetaToolAgent: Towards Generalizable Tool Usage in LLMs through Meta-Learning MetaToolAgent：通过元学习提升LLM在工具使用上的泛化能力 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🔗⭐
10 Distilling Time Series Foundation Models for Efficient Forecasting 提出DistilTS，用于高效蒸馏时序基础模型以实现高效预测。 distillation foundation model ✅
11 On the Relation of State Space Models and Hidden Markov Models 统一框架对比隐马尔可夫模型与状态空间模型，桥接控制理论、概率建模与深度学习。 Mamba SSM state space model
12 Balancing Classification and Calibration Performance in Decision-Making LLMs via Calibration Aware Reinforcement Learning 提出校准感知强化学习，平衡决策LLM的分类性能与校准置信度 reinforcement learning large language model
13 Analysis of Long Range Dependency Understanding in State Space Models 针对S4D模型，提出首个基于核解释性的长程依赖理解分析方法，应用于源代码漏洞检测。 SSM state space model
14 Training instability in deep learning follows low-dimensional dynamical principles 提出统一的动态视角，研究深度学习训练过程中的不稳定性问题 reinforcement learning large language model
15 Recursive Meta-Distillation: An Axiomatic Framework for Iterative Knowledge Refinement 提出递归元蒸馏框架，为迭代知识精炼提供公理化理论基础。 distillation
16 Knowledge-Integrated Representation Learning for Crypto Anomaly Detection under Extreme Label Scarcity; Relational Domain-Logic Integration with Retrieval-Grounded Context and Path-Level Explanations 提出RDLI框架，解决加密货币异常检测中标签稀缺和对抗性攻击问题 representation learning
17 Distribution-Centric Policy Optimization Dominates Exploration-Exploitation Trade-off 提出分布中心策略优化(DCPO)，解决LLM强化学习中探索-利用难题 reinforcement learning large language model ✅
18 Decoding Rewards in Competitive Games: Inverse Game Theory with Entropy Regularization 提出基于熵正则化的逆博弈论框架，用于竞争博弈中的奖励函数重构。 reinforcement learning inverse reinforcement learning

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#题目一句话要点标签🔗⭐
19 A Boolean Function-Theoretic Framework for Expressivity in GNNs with Applications to Fair Graph Mining 提出基于布尔函数理论的GNN表达性框架，应用于公平图挖掘。 OMOMO

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
20 TrustEnergy: A Unified Framework for Accurate and Reliable User-level Energy Usage Prediction TrustEnergy：用于精准可靠用户级能源使用预测的统一框架 spatiotemporal

⬅️ 返回 cs.LG 首页 · 🏠 返回主页