Running 596 Scaling test-time compute 📈 596 Run advanced search strategies to boost LLM problem solving
OpenSeeker: Democratizing Frontier Search Agents by Fully Open-Sourcing Training Data Paper • 2603.15594 • Published Mar 16 • 149
Toward Ultra-Long-Horizon Agentic Science: Cognitive Accumulation for Machine Learning Engineering Paper • 2601.10402 • Published Jan 15 • 37