papers.joyehuang.me

Paper Recommendations

Daily AI-powered curation from arXiv and Hugging Face Daily Papers. Last updated May 23, 2026 · 10:17 UTC.

Archive joyehuang.me Source
90
Candidates
5
Top Picks
0
Blockbusters
17
Max Score
Today's Picks

Editor's Rationale

🤖 今日推荐（Agent 开发工程师精选） 本周聚焦 Agent 架构的自我进化、多 Agent 通信安全、模型编排策略和 MCP 工具基础设施——四篇各有侧重的实用论文。 --- **1. MOSS: Self-Evolution through Source-Level Rewriting in Autonomous Agent Systems** ⭐ 最推荐 现有 Agent 系统的自我进化只停留在 prompt/skill 等文本层面，无法触及真正的代码级缺陷（路由、hook 顺序、状态不变量）。MOSS 让 Agent 直接在源码层自我重写——从生产 failure 证据出发，经多阶段流水线生成代码修改、用 ephemeral 容器验证、再通过用户同意的容器热替换部署。在 OpenClaw 上单次循环就将 mean grader score 从 0.25 提升到 0.61。对做 Agent 架构的读者来说，这是“Agent 能否真正自我修复”方向的关键突破。 **2. Maestro: Reinforcement Learning to Orchestrate Hierarchical Model-Skill Ensembles** 不再试图把所有能力塞进一个 monolithic 模型。Maestro 训练一个轻量 4B 参数策略，动态编排多个 frozen 专家模型和技能库，通过 RL 优化选择策略。在 10 个多模态 benchmark 上以 70.1% 平均准确率超越 GPT-5（69.3%）和 Gemini-2.5-Pro（68.7%），且学到的编排策略能泛化到未见过的模型和技能。对关心 Multi-Agent 编排和 Tool Use 架构的读者极有价值。 **3. LCGuard: Latent Communication Guard for Safe KV Sharing in Multi-Agent Systems** 多 Agent 系统通过 KV cache 做 latent communication 能保留更丰富的语义信息，但也带来了安全隐患——敏感内容可在 agent 间无声传播。LCGuard 提出在共享 KV 前做 representation-level 变换，通过对抗训练平衡任务性能和信息泄露风险。对构建多 Agent 系统的团队来说，这是个不可忽视的安全基础设施层。 **4. HarnessAPI: A Skill-First Framework for Unified Streaming APIs and MCP Tools** 每一个 Python 函数部署为 LLM tool 都要同时维护 HTTP endpoint 和 MCP tool 两份代码。HarnessAPI 以 typed skill folder 为单一事实源，自动生成 SSE streaming HTTP 接口 + OpenAPI UI + MCP tool，减少 74% 框架层样板代码。如果你在用或打算用 MCP 协议，这篇值得一读。

Top Picks

5 papers
#01 17

ChronoMedKG: A Temporally-Grounded Biomedical Knowledge Graph and Benchmark for Clinical Reasoning

Md Shamim Ahmed, Farzaneh Firoozbakht, Lukas Galke Poech et al.
cs.CL
Biomedical knowledge graphs (KGs) treat disease associations as static facts, but temporal information is crucial for clinical reasoning, e.g., a symptom diagnostic of one disease at age 3 may imply a different disease at age 13. Existing KGs such as PrimeKG, Hetionet, and iKraph…
agentmulti-agentreasoningautonomousRAGretrieval-augmented
#02 16

ClinSeekAgent: Automating Multimodal Evidence Seeking for Agentic Clinical Reasoning

Juncheng Wu, Letian Zhang, Yuhan Wang et al.
Large language models (LLMs) and agentic systems have shown promise for clinical decision support, but existing works largely assume that evidence has already been curated and handed to the model. Real-world clinical workflows instead require agents to actively seek, iteratively …
agentagenticreasoningworkflowRAGPPO
#03 13

Boiling the Frog: A Multi-Turn Benchmark for Agentic Safety

Piercosma Bisconti, Matteo Prandi, Federico Pierucci et al.
cs.CL
Background. Traditional safety benchmarks for language models evaluate generated text: whether a model outputs toxic language, reproduces bias, or follows harmful instructions. When models are deployed as agents, the safety-relevant object shifts from what the system says to what…
agentagenticRAGevaluationbenchmark
#04 13

Agentic CLEAR: Automating Multi-Level Evaluation of LLM Agents

Asaf Yehudai, Lilach Eden, Michal Shmueli-Scheuer
cs.CL, cs.AI
Agentic systems are becoming more capable: agents define strategies, take actions, and interact with different environments. This autonomy poses serious challenges for overseeing and assessing agent behavior. Most current tools are limited, focusing on observability with basic ev…
agentagenticevaluationbenchmarkalignment
#05 13

Lean Refactor: Multi-Objective Controllable Proof Optimization via Agentic Strategy Search

Jialin Lu, Soonho Kong, Rodrigo Stehling et al.
We present Lean Refactor, a plug-and-play retrieval-augmented agentic framework for multi-objective, controllable, and version-robust refactoring of Lean proofs. LLM-generated proofs are notoriously correct-but-verbose and brittle across library versions, yet existing refactoring…
agentagenticRAGretrieval-augmentedfine-tuningbenchmark

All Candidates

90 papers
ScorePaperAuthorsCategory
12 MOSS: Self-Evolution through Source-Level Rewriting in Autonomous Agent Systems Qianshu Cai, Yonggang Zhang et al. cs.AI, cs.LG
12 LCGuard: Latent Communication Guard for Safe KV Sharing in Multi-Agent Systems Sadia Asif, Mohammad Mohammadi Amiri et al. cs.AI, cs.ET, cs.LG, cs.MA
12 Superhuman Safe and Agile Racing through Multi-Agent Reinforcement Learning Ismail Geles, Leonard Bauersfeld et al. cs.RO, cs.AI, cs.LG, cs.MA
11 One Sentence, One Drama: Personalized Short-Form Drama Generation via Multi-Agent Systems Yufei Shi, Weilong Yan et al.
10 LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning Yifan Dai, Zhenhua Wu et al.
10 Maestro: Reinforcement Learning to Orchestrate Hierarchical Model-Skill Ensembles Jinyang Wu, Guocheng Zhai et al.
10 π-Bench: Evaluating Proactive Personal Assistant Agents in Long-Horizon Workflows Haoran Zhang, Luxin Xu et al.
10 TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks Zhaoyang Chu, Jiarui Hu et al.
9 Bernini: Latent Semantic Planning for Video Diffusion Bernini Team, Chenchen Liu et al.
9 Spreadsheet-RL: Advancing Large Language Model Agents on Realistic Spreadsheet Tasks via Reinforcement Learning Banghao Chi, Yining Xie et al.
8 Rule2DRC: Benchmarking LLM Agents for DRC Script Synthesis with Execution-Guided Test Generation Jinuk Kim, Junsoo Byun et al.
8 Perception or Prejudice: Can MLLMs Go Beyond First Impressions of Personality? Caixin Kang, Tianyu Yan et al.
7 Advancing Mathematics Research with AI-Driven Formal Proof Search George Tsoukalas, Anton Kovsharov et al. cs.AI
7 Towards a General Intelligence and Interface for Wearable Health Data Girish Narayanswamy, Maxwell A. Xu et al. cs.AI
7 HarnessAPI: A Skill-First Framework for Unified Streaming APIs and MCP Tools Edwin Jose cs.AI, cs.SE
7 TransitLM: A Large-Scale Dataset and Benchmark for Map-Free Transit Route Generation Hanyu Guo, Jiedong Yang et al.
7 GenEvolve: Self-Evolving Image Generation Agents via Tool-Orchestrated Visual Experience Distillation Sixiang Chen, Zhaohu Xing et al.
6 Beyond Acoustic Emotion Recognition: Multimodal Pathos Analysis in Political Speech Using LLM-Based and Acoustic Emotion Models Juergen Dietrich cs.AI, cs.CL, cs.HC, cs.SD, eess.AS
6 OmniPro: A Comprehensive Benchmark for Omni-Proactive Streaming Video Understanding Ruixiang Zhao, Jie Yang et al.
6 Sensor2Sensor: Cross-Embodiment Sensor Conversion for Autonomous Driving Jiahao Wang, Bo Sun et al.
5 DeltaBox: Scaling Stateful AI Agents with Millisecond-Level Sandbox Checkpoint/Rollback Yunpeng Dong, Jingkai He et al. cs.OS, cs.AI
5 AMEL: Accumulated Message Effects on LLM Judgments Sid-ali Temkit cs.AI, cs.CL, cs.LG
5 Two is better than one: A Collapse-free Multi-Reward RLIF Training Framework Shourov Joarder, Diganta Sikdar et al. cs.LG, cs.CL
5 SpaceDG: Benchmarking Spatial Intelligence under Visual Degradation Xiaolong Zhou, Yifei Liu et al.
4 The Matching Principle: A Geometric Theory of Loss Functions for Nuisance-Robust Representation Learning Vishal Rajput cs.LG, cs.AI, stat.ML
4 Moral Semantics Survive Machine Translation: Cross-Lingual Evidence from Moral Foundations Corpora Maciej Skorski cs.CL, cs.AI
4 Forecasting Scientific Progress with Artificial Intelligence Sean Wu, Pan Lu et al.
4 Diversed Model Discovery via Structured Table Discovery Zhengyuan Dong, Renée J. Miller
3 Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention Ali Hatamizadeh, Yejin Choi et al. cs.AI
3 Deep Reinforcement Learning for Flexible Job Shop Scheduling with Random Job Arrivals Yu Tang, Muhammad Zakwan et al. cs.AI, math.OC
3 The Distillation Game: Adaptive Attacks & Efficient Defenses Youssef Allouah, Mahdi Haghifam et al. cs.LG, cs.AI
3 Evaluating Commercial AI Chatbots as News Intermediaries Mirac Suzgun, Emily Shen et al. cs.CL
3 AutoRubric-T2I: Robust Rule-Based Reward Model for Text-to-Image Alignment Kuei-Chun Kao, Daixuan Huo et al.
3 Live Music Diffusion Models: Efficient Fine-Tuning and Post-Training of Interactive Diffusion Music Generators Zachary Novack, Stephen Brade et al.
3 FashionLens: Toward Versatile Fashion Image Retrieval via Task-Adaptive Learning Haokun Wen, Xuemeng Song et al.
3 Forecasting Downstream Performance of LLMs With Proxy Metrics Arkil Patel, Siva Reddy et al.
3 DecQ: Detail-Condensing Queries for Enhanced Reconstruction and Generation in Representation Autoencoders Tianhang Wang, Yitong Chen et al.
3 Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps Yanke Zhou, Yiduo Li et al.
2 MambaGaze: Bidirectional Mamba with Explicit Missing Data Modeling for Cognitive Load Assessment from Eye-Gaze Tracking Data Amir Mousavi, Mohammad Sadegh Sirjani et al. cs.LG, cs.AI, cs.HC
2 CogAdapt: Transferring Clinical ECG Foundation Models to Wearable Cognitive Load Assessment via Lead Adaptation Amir Mousavi, Mohammad Sadegh Sirjani et al. cs.LG, cs.AI, cs.HC
2 Reducing Political Manipulation with Consistency Training Long Phan, Devin Kim et al. cs.CL, cs.AI
2 Understanding Data Temporality Impact on Large Language Models Pre-training Pilchen Hippolyte, Fabre Romain et al. cs.CL, cs.AI
2 Cyber-Physical Anomaly Detection in IoT-Enabled Smart Grids Using Machine Learning and Metaheuristic Feature Optimization Adis Alihodžić, Eva Tuba et al. cs.LG, cs.AI
2 Remember to be Curious: Episodic Context and Persistent Worlds for 3D Exploration Lily Goli, Justin Kerr et al. cs.LG
2 Lumberjack: Better Differentially Private Random Forests through Heavy Hitter Detection in Trees Christian Janos Lebeda, David Erb et al. cs.LG, cs.DS
2 AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the Wild Baiyu Chen, Zechen Li et al. cs.CV, cs.AI, cs.CL, cs.HC
2 Self-Policy Distillation via Capability-Selective Subspace Projection Guangya Hao, Yitong Shang et al. cs.CL
2 More Context, Larger Models, or Moral Knowledge? A Systematic Study of Schwartz Value Detection in Political Texts Víctor Yeste, Paolo Rosso cs.CL, cs.AI, cs.LG
2 Segment Anything with Motion, Geometry, and Semantic Adaptation for Complex Nonlinear Visual Object Tracking Deyi Zhu, Yuji Wang et al.
2 DelTA: Discriminative Token Credit Assignment for Reinforcement Learning from Verifiable Rewards Kaiyi Zhang, Wei Wu et al.
2 SceneAligner: 3D-Grounded Floorplan Localization in the Wild Junhyeong Cho, Ruojin Cai et al.
2 Training Large Language Models to Predict Clinical Events Benjamin Turtel, Paul Wilczewski et al.
2 From Reasoning Chains to Verifiable Subproblems: Curriculum Reinforcement Learning Enables Credit Assignment for LLM Reasoning Xitai Jiang, Zihan Tang et al.
2 WorldKV: Efficient World Memory with World Retrieval and Compression Jung Yi, Minjae Kim et al.
1 Proxy-Based Approximation of Shapley and Banzhaf Interactions Santo M. A. R. Thies, Hubert Baniecki et al. cs.LG, cs.AI, stat.ML
1 Integrable Elasticity via Neural Demand Potentials Carlos Heredia, Daniel Roncel cs.LG
1 Plug-in Losses for Evidential Deep Learning: A Simplified Framework for Uncertainty Estimation that Includes the Softmax Classifier Berk Hayta, Hannah Laus et al. cs.LG, eess.AS, stat.ML
1 SeqLoRA: Bilevel Orthogonal Adaptation for Continual Multi-Concept Generation Javad Parsa, Enis Simsar et al. cs.LG
1 Ternary Decision Trees with Locally-Adaptive Uncertainty Zones William Smits cs.LG
1 The Double Dilemma in Multi-Task Radiology Report Generation: A Gradient Dynamics Analysis and Solution Erjian Zhang, Yatong Hao et al. cs.LG, cs.CL, cs.CV
1 SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild Xuyi Hu, Jin Lyu et al.
1 Platonic Representations in the Human Brain: Unsupervised Recovery of Universal Geometry Pablo Marcos-Manchón, Rishi Jha et al.
1 Disentangling Sampling from Training Budget in Class-Imbalanced CT Body Composition Segmentation Iason Skylitsis, Dimitrios Karkalousos et al.
1 SEGA: Spectral-Energy Guided Attention for Resolution Extrapolation in Diffusion Transformers Javad Rajabi, Kimia Shaban et al.
1 "I didn't Make the Micro Decisions": Measuring, Inducing, and Exposing Goal-Level AI Contributions in Collaboration Eunsu Kim, Jessica R. Mindel et al.
1 Minimalist Visual Inertial Odometry Francesco Pasti, Jeremy Klotz et al.
1 FlowLong: Inference-time Long Video Generation via Manifold-constrained Tweedie Matching Jangho Park, Geon Yeong Park et al.
1 KVServe: Service-Aware KV Cache Compression for Communication-Efficient Disaggregated LLM Serving Zedong Liu, Xinyang Ma et al.
1 Swift Sampling: Selecting Temporal Surprises via Taylor Series Dahye Kim, Bhuvan Sachdeva et al.
0 Finite-Particle Convergence Rates for Conservative and Non-Conservative Drifting Models Krishnakumar Balasubramanian stat.ML, cs.AI, cs.LG, math.ST
0 SDPM: Survival Diffusion Probabilistic Model for Continuous-Time Survival Analysis Stanislav R. Kirpichenko, Andrei V. Konstantinov et al. cs.LG, cs.AI, stat.CO, stat.ML
0 Tokenisation via Convex Relaxations Jan Tempus, Philip Whittington et al. cs.CL, cs.LG
0 FAME: Failure-Aware Mixture-of-Experts for Message-Level Log Anomaly Detection Huanchi Wang, Zihang Huang et al. cs.SE, cs.LG
0 Uniform Diffusion Models Revisited: Leave-One-Out Denoiser and Absorbing State Reformulation Samson Gourevitch, Yazid Janati et al. cs.LG, stat.ML
0 Tokenization with Split Trees Craig W. Schmidt, Michael Krumdick et al. cs.CL
0 Seeing the Poem: Image-Semantic Detection of AI-Generated Modern Chinese Poetry with MLLMs Shanshan Wang, Fengying Ye et al. cs.CL, cs.CV
0 Whose Voice Counts? Mapping Stakeholder Perspectives on AI Through Public Submissions to the U.S. Government Alina Karakanta, Alex Christiansen et al. cs.CL
0 Chinese sensorimotor and embodiment norms for 3,000 lexicalized concepts Jing Chen, Gábor Parti et al. cs.CL
0 LoREnc: Low-Rank Encryption for Securing Foundation Models and LoRA Adapters Beomjin Ahn, Jungmin Kwon et al.
0 Q-ARVD: Quantizing Autoregressive Video Diffusion Models Siao Tang, Xinyin Ma et al.
0 PhysX-Omni: Unified Simulation-Ready Physical 3D Generation for Rigid, Deformable, and Articulated Objects Ziang Cao, Yinghao Liu et al.
0 Same Architecture, Different Capacity: Optimizer-Induced Spectral Scaling Laws Nandan Kumar Jha, Brandon Reagen
0 RiT: Vanilla Diffusion Transformers Suffice in Representation Space Le Zhang, Ning Mang et al.
-1 Vector Policy Optimization: Training for Diversity Improves Test-Time Search Ryan Bahlous-Boldi, Isha Puri et al. cs.LG, cs.AI, cs.CL, cs.NE
-4 Unsupervised Process Reward Models Artyom Gadetsky, Maxim Kodryan et al.