JANUS: Enhancing Long-Horizon Autonomous Research with Co-Evolving World Models

arXiv:2610.13194 · cs.AI, cs.CL, cs.MA · Submitted 2026-10-09 · Read on arXiv

cs.AI, cs.CL, cs.MA

Submitted: 2026-10-09

Updated: 2026-10-09

Code: https://github.com/bilemond/JANUS

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• World Model — 1× in this paper · explained in 7 episodes

Sources

• K-Search: LLM Kernel Generation via Co-Evolving Intrinsic World Model
• MARS: Modular Agent with Reflective Search for Automated AI Research
• MLEvolve: A Self-Evolving Framework for Automated Machine Learning Algorithm Discovery
• TinyStories: How Small Can Language Models Be and Still Speak Coherent English?
• MLReplicate: Benchmarking Autonomous Research Systems for Machine Learning Reproducibility
• Mathematical exploration and discovery at scale
• Is Your LLM Secretly a World Model of the Internet? Model-Based Planning for Web Agents
• AutoKernel: Autonomous GPU Kernel Optimization via Iterative Agent-Driven Search
• CodeScientist: End-to-End Semi-Automated Scientific Discovery with Code-based Experimentation
• AIDE: AI-Driven Exploration in the Space of Code
• Toward Generalist Autonomous Research via Hypothesis-Tree Refinement
• AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration
• MindFlow: Mind Supernet Powered Thinking Flows for Research Idea Innovation
• ML-Master: Towards AI-for-AI via Integration of Exploration and Reasoning
• The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
• MLS-Bench: A Holistic and Rigorous Assessment of AI Systems on Building Better AI
• EvoScientist: Towards Multi-Agent Evolving AI Scientists for End-to-End Scientific Discovery
• ScientistOne: Towards Human-Level Autonomous Research via Chain-of-Evidence
• Kosmos: An AI Scientist for Autonomous Discovery
• AlphaEvolve: A coding agent for scientific and algorithmic discovery

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems