A living map of AI research

See what AI is learning next.

Follow the newest papers, explore research as it forms into topics, and spot shifts in attention before they become obvious.

Browse the latest papers →Ask the research agent See research trends

Latest paper indexedSeptember 15Updated continuously from arXiv

Published papers58.8K

Living topics3,550

Fresh from the index

New research, without the noise

View all 58.8K papers →

Agentic Societies Need a Social Harness
Sep 15, 2026Tapan Chugh, Vidushi Singh, Krish Jain +2Agent Harness Complex Societies
Modality-Autoregressive World-Action Models
Sep 15, 2026Adam Hung, Bardienus P. Duisterhof, Deva Ramanan +1Action-Conditioned World Modeling
ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents
Sep 15, 2026Shuhan Xue, Jianyuan Zhong, Ziyuan Nan +10
PhysStream: Streaming Physics-Grounded Video Generation with Structured Scene Memory and Fine-Grained Motion Control
Sep 15, 2026Chuhao Chen, Peter Wonka, Chaoyang Wang +4Controllable High-Quality Human Video Generation Human Motion Generation
When Should LLMs Abstain? Chain-of-Self-Questioning for Selective Risk Control
Sep 15, 2026Ali ŞenolLarge Language Model Reliability
What Breaks Under Pruning in Smart Homes, and When? Evaluating LLM Degradation Across Architectures and Task Complexity
Sep 15, 2026Congjing Zhang, Vashishtha Patil, Henning Lange +1

Sep 15, 2026 · cs.MAJ/K move · Enter open · S save

Agentic Societies Need a Social Harness

Tapan Chugh, Vidushi Singh, Krish Jain, Arvind Krishnamurthy+1

An agentic society is a collection of AI agents that coordinate autonomously across trust boundaries, on behalf of different principals whose objectives may only partially align. We show experimentally that in agentic societies even honest, competent agents often fail to reach satisfactory outcomes with existing harnesses and messaging primitives, and that faulty or malicious agents can stall collaboration, influence outcomes, and pursue other harmful goals by exploiting vulnerabilities in communication (``speech''). We argue that agentic societies need a \emph{social harness} for inter-agent interactions, in addition to each agent's \emph{personal harness}, which manages its private context and communication with its principal. We propose a layered architecture for social harnesses which (i) prevents classes of failures outright, (ii) enables agents to detect invalid messages at runtime, and (iii) supports post-facto investigation and consequences, and highlight directions for future research to realize these capabilities.

Agent Harness Complex Societies

Open paper Ask about it PDF

The emerging map

Browse research fields

Fields are broad, emergent lenses above the more focused topics. They change as the research landscape changes.

Research field