This page shows the most relevant public items for al:agentic-ai-2025, ranked by trend activity and review signal. Use weekly for fast changes, monthly for more stable patterns, and all-time for evergreen picks.
Paper • Aug 8, 2025 • arXiv • Jian Lu, Percy Liang, Chelsea Finn
While scaling laws for single-model LLMs are well established, the relationship between the number of collaborating agents and overall system performance remains poorly understood. This paper inves...
Paper • Dec 22, 2025 • arXiv • Percy Liang, Thomas K. V., Eleanor Rigby
Aligning multi-agent systems via traditional human feedback is intractable due to the sheer volume and speed of agent-to-agent interactions. We introduce a novel alignment framework utilizing Inver...
Paper • Nov 18, 2025 • arXiv • Ling Chen, David Wu, Yong Zhang
The migration of legacy codebases (e.g., COBOL, Fortran) to modern architectures is a notoriously slow, error-prone, and expensive human endeavor. We introduce 'AutoMigrate,' a persistent, multi-ag...
Paper • Jul 28, 2025 • arXiv • Sarah Jenkins, Esq., Wei Chen, Marcus T. Holden
The legal discovery process involves manually cross-referencing millions of pages of depositions and emails to find logical contradictions. Standard search algorithms fail to connect multi-document...
Paper • Oct 14, 2025 • arXiv • J. L. Martinez, Susan Athey, Alessandro V. Rossi
As the internet transitions to a machine-to-machine economy, Agentic AI systems frequently need to outsource sub-tasks (e.g., specialized data retrieval, proprietary compute) to external agents. Ho...
Paper • May 22, 2025 • arXiv • Wei Lin, Chen Wei, Yujin Han
One of the most persistent failure modes in autonomous Agentic AI is the 'Infinite Loop'—where an agent becomes stuck continuously re-executing a failed API call or repeating a flawed logical deduc...
Paper • Nov 26, 2025 • arXiv • Ling Chen, Song Han, Yonggan Fu
The deployment of personal Agentic AI assistants is severely hindered by the privacy risks associated with uploading sensitive personal data (e.g., financial records, private messages) to centraliz...
Paper • Jun 30, 2025 • arXiv • Alexander C. K., Wei Chen, Torsten Hoefler
Current Agentic AI frameworks rely on high-level Python libraries that awkwardly manage memory, context windows, and tool execution at the application layer. We propose Agent-OS, a novel operating ...
Paper • Oct 21, 2025 • arXiv • Thomas K. V., Yiming Zhang, Nils Bergmann
The integration of decentralized renewable energy sources and the massive power demands of AI datacenters have created unprecedented volatility in modern power grids. We introduce 'Energy Swarm', a...
Paper • Aug 12, 2025 • arXiv • Barr Moses, Jian Lu, Sophia M.
Data engineering teams spend the majority of their time fixing broken Extract, Transform, Load (ETL) pipelines caused by upstream schema changes or data drift. We introduce ADEPT (Autonomous Data E...
Paper • Apr 16, 2025 • arXiv • Yannick LeCun, Antoine Bordes, Elena Rossi, Marcus Thorne
The propensity of large language models to hallucinate facts severely limits their deployment in autonomous agentic workflows where verification is costly. This paper introduces 'Agentic Debate,' a...
Paper • Dec 1, 2025 • arXiv • Susan Athey, Michael I. Jordan
As Agentic AI systems are deployed to autonomously negotiate contracts, manage supply chains, and execute high-frequency trades, they form a complex, decentralized digital economy. This paper appli...
Paper • Jul 22, 2025 • arXiv • Wei Chen, Yujin Han, Qingwen Bu
Current Agentic AI systems are constrained by the predefined set of tools provided by developers. We introduce AnyTool, a framework that grants agents the autonomy to dynamically generate, test, an...
Paper • Apr 18, 2025 • arXiv • Maciej Besta, Nils Blach, Torsten Hoefler
While Chain-of-Thought and Tree-of-Thoughts prompting greatly enhance LLM reasoning, they strictly enforce linear or hierarchical cognitive paths. We introduce Graph of Thoughts (GoT), a novel cogn...
Paper • Sep 10, 2025 • arXiv • Yao Mu, Tianyu Zheng, Percy Liang, Dan Hendrycks
As multi-agent swarms are deployed in open-ended web environments, standard static benchmarks fail to capture their vulnerability to dynamic threats and deceptive actors. We introduce AgentBench-20...
Paper • Aug 14, 2025 • arXiv • Sergey Levine, Chelsea Finn, Ankur Handa, Dieter Fox
While Agentic AI has mastered digital environments through API integrations, transferring this autonomous orchestration to the physical world remains a significant challenge. This paper presents a ...
Paper • Sep 14, 2025 • arXiv • Michael C. Brooks, Anita Desai, Francis Nguyen
As Agentic AI systems assume more responsibility in high-stakes domains, calibrating human trust remains a critical challenge. Under-trust leads to inefficient micromanagement, while over-trust res...
Paper • Aug 11, 2025 • arXiv • Isabella R. Rossi, Kenji Sato, Nils Bergmann
The discovery of novel chemical compounds is traditionally hindered by the slow, iterative process of manual hypothesis generation and bench testing. We present ChemAgent, an Agentic AI framework t...
Paper • Jul 15, 2025 • arXiv • Kevin Zhu, Lara Croft, Julian Bao
Evaluating the long-horizon planning and adaptability of Agentic AI in the real world is fraught with safety and cost limitations. We establish Minecraft as the premier sandbox for open-ended agent...
Paper • Jun 21, 2025 • arXiv • Thomas K. V., Eleanor Rigby, M. Hasan
Traditional AI alignment techniques like RLHF are insufficient for Agentic AI, as autonomous systems frequently encounter novel edge cases in open-ended environments that were absent from their tra...
It ranks public content for al:agentic-ai-2025 using recent discussion, review, and engagement signals so you can triage faster. This guidance is specific to al:agentic-ai-2025 topic page on Attendemia and is written so it still makes sense without reading other sections on the page.
Use weekly for fast-moving updates, monthly for stable trend confirmation, and all-time for evergreen references. This guidance is specific to al:agentic-ai-2025 topic page on Attendemia and is written so it still makes sense without reading other sections on the page.
Use the linked entities section to jump to labs, companies, and experts connected to this topic and explore their timelines. This guidance is specific to al:agentic-ai-2025 topic page on Attendemia and is written so it still makes sense without reading other sections on the page.
Yes. Use the follow button on this page to subscribe and track new high-signal activity. This guidance is specific to al:agentic-ai-2025 topic page on Attendemia and is written so it still makes sense without reading other sections on the page.