Discovery Foundation Models: Toward Open-Ended Discovery Intelligence
cs.CL
Submitted: 2026-09-14
Updated: 2026-09-14
Comments: Website: https://phai-labs.com/collaborate/, Code: https://github.com/Gen-Verse/DFM-Plans
Code: https://github.com/Gen-Verse/DFM-Plans
Project page: https://phai-labs.com/collaborate
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- On the Opportunities and Risks of Foundation Models
- ResearchGym: Evaluating Language Model Agents on Real-World AI Research
- GenEnv: Difficulty-Aligned Co-Evolution Between LLM Agents and Environment Simulators
- AI scientists produce results without reasoning scientifically
- Language agents achieve superhuman synthesis of scientific knowledge
- Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters
- Evaluating Large Language Models in Scientific Discovery
- ReasonFlux: Hierarchical LLM Reasoning via Scaling Thought Templates
- Recursive Experiential-Working Memory Evolution for Long-Horizon Agent Harnesses
- Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering