cs.LG（2026-10-21）

📊 共 35 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (20 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (13) 支柱八：物理动画 (Physics-based Animation) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (20 篇)

#题目一句话要点标签🔗⭐
1 YingLan: A Multimodal Ocean Foundation Model for Unified Representation Learning from Heterogeneous Observations 提出YingLan以解决多模态海洋建模中的异构观察问题 representation learning spatiotemporal foundation model
2 Value-Rectified Distillation for Flow-based Offline Reinforcement Learning 提出价值修正蒸馏以解决离线强化学习中的OOD模式平均问题 reinforcement learning offline reinforcement learning flow matching
3 How Far Can Foundation Models Be Distilled for Efficient Jet Tagging? 通过知识蒸馏提升喷流标记效率的研究 distillation foundation model
4 Compositional neural networks for multimodal X-ray and neutron scattering inversion 提出模块化神经网络以解决多模态X射线与中子散射反演问题 flow matching multimodal
5 Causal world models under physical parameter shifts 提出因果世界模型以应对物理参数变化问题 world model world models
6 Flat-RLVR: Sharpness-Aware Reinforcement Learning with Verifiable Rewards for Reasoning Capacity in LLMs 提出Flat-RLVR以解决LLMs推理能力下降问题 reinforcement learning large language model
7 Differentially Private Sparse Reward Estimation with Preference Feedback 提出差分隐私稀疏奖励估计方法以解决用户偏好数据泄露问题 reinforcement learning RLHF large language model
8 EPIC: Rethinking Trust Regions for Autoregressive LLM Policies via Exact-KL Prefix Importance-Ratio Clipping 提出EPIC以解决自回归LLM策略中的信任区域问题 reinforcement learning PPO large language model ✅
9 CORD: Conflict-Aware Optimization for Joint RL and Distillation in Agents 提出CORD框架以解决联合强化学习与蒸馏训练中的冲突问题 reinforcement learning distillation
10 Stochastic Pruning of the DPO Computation Graph: 1.4x Faster Training and Lower Peak Memory 提出随机剪枝框架以加速DPO训练并降低内存峰值 DPO direct preference optimization
11 Are Joint-Embedding Predictive Objectives Effective for Longitudinal EHRs? 探讨联合嵌入预测目标在纵向电子健康记录中的有效性 JEPA Joint-Embedding Predictive Architecture joint-embedding predictive architecture
12 Dissimilarity-Based Markov Decision Processes for Stabilizing State Distributions 提出基于不相似度的马尔可夫决策过程以稳定状态分布 reinforcement learning policy learning PPO
13 Near Minimax Rate for Quantile Distributional Reinforcement Learning 提出量化分布强化学习的新方法以提高策略评估精度 reinforcement learning
14 Shortcut-Resistant CAM Distillation for Long-Tailed Recognition 提出Shortcut-Resistant CAM蒸馏以解决长尾识别问题 distillation ✅
15 Harness Scaling for Generalizable Agentic Reinforcement Learning 提出Harness Scaling以解决强化学习中的通用性问题 reinforcement learning
16 BIRD-RL: Scaling Agentic Reinforcement Learning over Stateful Data-Centric Environments 提出BIRD-RL以解决状态数据中心环境中的强化学习效率问题 reinforcement learning
17 Parallelizable Probabilistic Latent Reasoning for Diverse Reasoning Trajectories 提出PLRM以解决多样化推理轨迹建模问题 JEPA chain-of-thought
18 DISRL: Dual-Scale Incremental Semi-Supervised Reinforcement Learning for Dynamic Community Tracking 提出DISRL以解决动态社区跟踪中的监督不确定性与错误累积问题 reinforcement learning
19 MA-UITD: Offline Multi-Agent Reward Inference for Characterizing Continuous Social Behavior 提出MA-UITD以解决多智能体奖励推断问题 reinforcement learning imitation learning inverse reinforcement learning
20 Distilling Multi-Teacher Scoring Principles for Unsupervised Time Series Anomaly Detection 提出AnoMix以解决无监督时间序列异常检测中的多教师评分问题 distillation foundation model

🔬 支柱九：具身大模型 (Embodied Foundation Models) (13 篇)

#题目一句话要点标签🔗⭐
21 Monitorability Is Trainable: Post-Training for Chain-of-Thought Monitorability 提出无监督后训练方法以提升链式思维监控能力 chain-of-thought
22 Leveraging Time-Series Foundation Models for Financial Crime Detection 提出时间序列基础模型以解决金融犯罪检测问题 foundation model
23 Reliability-Aware Probabilistic Fusion for Multimodal Learning with Incomplete Evidence 提出可靠性感知的高斯专家融合方法以解决多模态学习中的证据缺失问题 multimodal
24 Privacy in Context Is Being Overlooked in the Age of General-Purpose LLMs 提出针对通用LLM的隐私保护框架以解决隐私上下文问题 large language model
25 Recursive Code Evolution for Task-Oriented Semantic Communications 提出ComRSI框架以优化任务导向的语义通信 large language model
26 WRDO: Weighted Refusal Direction Optimization in Aligned Language Models 提出WRDO以增强对有害请求的拒绝能力 large language model
27 Residual-Based Causal Discovery via Pretrained In-Context Quantile Regression 提出Loscar以解决LSNM中的残差因果发现问题 foundation model
28 Rethinking Agentic Model Orchestration for Time Series Forecasting 提出FairAMO以解决时间序列预测中的模型协调问题 foundation model
29 Nominate-then-Adjudicate: LLM-Assisted One-Pass and Per-Instance MIP Solver Tuning 提出Nominate-then-Adjudicate方法以优化混合整数规划求解器调优 foundation model
30 Benchmarking and Error Analysis of Deep Search and Deep Research LLM Agents 提出全面方法以诊断深度搜索和研究代理的中间错误 large language model
31 ExpEvolve: Structuring and Reusing Experience for Open-Ended Discovery 提出ExpEvolve以解决开放式科学发现中的经验重用问题 large language model
32 Approximate Solomonoff Induction: On the Robustness of Learned Universal Predictors 提出近似Solomonoff归纳以增强学习型通用预测器的鲁棒性 large language model
33 The Price of Choice in Selective Prediction 提出选择性预测中的成本分析以优化分类器性能 multimodal

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

#题目一句话要点标签🔗⭐
34 BrainCodes: A Lightweight EEG Foundation Model over Discrete Neural Codes 提出BrainCodes以解决EEG模型在资源受限设备上的应用问题 spatiotemporal foundation model

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

#题目一句话要点标签🔗⭐
35 Progressive Permutation Alignment with Temperature-Controlled Spectral Sharpening 提出渐进排列对齐方法以解决神经表示模型的对齐问题 NeRF neural radiance field

⬅️ 返回 cs.LG 首页 · 🏠 返回主页