cs.LG（2025-06-12）

📊 共 32 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (18 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (10 🔗2) 支柱四：生成式动作 (Generative Motion) (3 🔗1) 支柱一：机器人控制 (Robot Control) (1)

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

#题目一句话要点标签🔗⭐
1 FedNano: Toward Lightweight Federated Tuning for Pretrained Multimodal Large Language Models 提出FedNano以解决多模态大语言模型的轻量化联邦调优问题 large language model multimodal
2 Graph-MLLM: Harnessing Multimodal Large Language Models for Multimodal Graph Learning 提出Graph-MLLM以解决多模态图学习的评估与整合问题 large language model multimodal
3 Predictable Scale: Part II, Farseer: A Refined Scaling Law in Large Language Models 提出Farseer以解决大规模语言模型训练中的预测精度问题 large language model ✅
4 GUARD: Guided Unlearning and Retention via Data Attribution for Large Language Models 提出GUARD框架以解决大语言模型中的非意图遗忘问题 large language model
5 Foundation Models for Causal Inference via Prior-Data Fitted Networks 提出CausalFM以解决因果推断中的模型训练问题 foundation model
6 MNN-LLM: A Generic Inference Engine for Fast Large Language Model Deployment on Mobile Devices 提出MNN-LLM以解决移动设备上大语言模型推理速度慢的问题 large language model
7 Time-IMM: A Dataset and Benchmark for Irregular Multimodal Multivariate Time Series 提出Time-IMM数据集以解决不规则多模态多变量时间序列问题 multimodal ✅
8 EAGLE: Efficient Alignment of Generalized Latent Embeddings for Multimodal Survival Prediction with Interpretable Attribution Analysis 提出EAGLE以解决多模态癌症生存预测中的融合与可解释性问题 multimodal
9 Build the web for agents, not agents for the web 提出代理网络接口以解决现有网页代理适应性不足问题 large language model multimodal
10 Robustly Improving LLM Fairness in Realistic Settings via Interpretability 通过可解释性方法提升LLM在招聘中的公平性 large language model chain-of-thought
11 Data Shifts Hurt CoT: A Theoretical Study 研究数据偏移对链式思维的影响及其机制 large language model chain-of-thought
12 NoLoCo: No-all-reduce Low Communication Training Method for Large Models 提出NoLoCo以解决大模型训练中的通信瓶颈问题 large language model
13 Detecting High-Stakes Interactions with Activation Probes 提出激活探针以检测高风险交互问题 large language model
14 ConTextTab: A Semantics-Aware Tabular In-Context Learner 提出ConTextTab以解决表格数据语义理解不足的问题 large language model ✅
15 BugGen: A Self-Correcting Multi-Agent LLM Pipeline for Realistic RTL Bug Synthesis 提出BugGen以解决RTL调试效率低下的问题 large language model
16 Time To Impeach LLM-as-a-Judge: Programs are the Future of Evaluation 提出PAJAMA以解决LLM评估中的高成本与偏见问题 large language model
17 TreeLoRA: Efficient Continual Learning via Layer-Wise LoRAs Guided by a Hierarchical Gradient-Similarity Tree 提出TreeLoRA以解决高效持续学习问题 large language model
18 Provably Learning from Language Feedback 提出HELiX算法以解决语言反馈学习问题 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (10 篇)

#题目一句话要点标签🔗⭐
19 Can Time-Series Foundation Models Perform Building Energy Management Tasks? 提出时间序列基础模型以解决建筑能源管理任务的可扩展性问题 representation learning large language model foundation model
20 Collapsing Sequence-Level Data-Policy Coverage via Poisoning Attack in Offline Reinforcement Learning 提出序列级数据-策略覆盖崩溃攻击以解决离线强化学习安全问题 reinforcement learning offline RL offline reinforcement learning
21 Discovering Hierarchical Latent Capabilities of Language Models via Causal Representation Learning 提出因果表示学习框架以评估语言模型潜在能力 representation learning instruction following
22 Viability of Future Actions: Robust Safety in Reinforcement Learning via Entropy Regularization 通过熵正则化实现强化学习的鲁棒安全性 reinforcement learning reward shaping
23 Self-Adapting Language Models 提出自适应语言模型以解决静态模型适应性不足问题 reinforcement learning large language model ✅
24 Sequential-Parallel Duality in Prefix Scannable Models 提出前缀可扫描模型以实现高效序列推理 Mamba state space model linear attention
25 Logarithmic Smoothing for Adaptive PAC-Bayesian Off-Policy Learning 提出自适应PAC-Bayesian离线学习的新方法以提高数据质量 policy learning
26 Task Adaptation from Skills: Information Geometry, Disentanglement, and New Objectives for Unsupervised Reinforcement Learning 提出LSEPIN和WSEP以提升无监督强化学习的任务适应性 reinforcement learning
27 Hierarchical Adversarially-Resilient Multi-Agent Reinforcement Learning for Cyber-Physical Systems Security 提出HAMARL框架以增强网络物理系统的安全性 reinforcement learning
28 The Diffusion Duality 提出Duo方法以缩小离散扩散模型与自回归模型的性能差距 curriculum learning distillation ✅

🔬 支柱四：生成式动作 (Generative Motion) (3 篇)

#题目一句话要点标签🔗⭐
29 Execution Guided Line-by-Line Code Generation 提出执行引导的逐行代码生成方法以提升代码生成性能 classifier-free guidance large language model ✅
30 Constrained Diffusion Models for Synthesizing Representative Power Flow Datasets 提出约束扩散模型以合成代表性电力流数据集 physics-informed diffusion
31 What Exactly Does Guidance Do in Masked Discrete Diffusion Models 提出明确指导机制以优化掩蔽离散扩散模型的采样行为 classifier-free guidance

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
32 Assessing the Resilience of Automotive Intrusion Detection Systems to Adversarial Manipulation 评估汽车入侵检测系统对对抗性攻击的韧性 manipulation

⬅️ 返回 cs.LG 首页 · 🏠 返回主页