Asking like Socrates: Socrates helps VLMs understand remote sensing images Paper • 2511.22396 • Published Nov 27, 2025 • 5
Surgical Post-Training: Cutting Errors, Keeping Knowledge Paper • 2603.01683 • Published 3 days ago • 10
Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning Paper • 2510.19338 • Published Oct 22, 2025 • 115
Every Step Evolves: Scaling Reinforcement Learning for Trillion-Scale Thinking Model Paper • 2510.18855 • Published Oct 21, 2025 • 73
Learning on the Job: An Experience-Driven Self-Evolving Agent for Long-Horizon Tasks Paper • 2510.08002 • Published Oct 9, 2025 • 23