ShellElastic: Scalable Elastic Pretraining with Simultaneous Maximal Updates and Hyperparameter Transfer Paper • 2610.15124 • Published 7 days ago • 2
ShellElastic: Scalable Elastic Pretraining with Simultaneous Maximal Updates and Hyperparameter Transfer Paper • 2610.15124 • Published 7 days ago • 2
LED: LLM Enhanced Open-Vocabulary Object Detection without Human Curated Data Generation Paper • 2503.13794 • Published Mar 18, 2025
Latent Reward Steering: An Adaptive Inference-Time Framework that Implicitly Promotes Cognitive Behaviors in Reasoning LLMs Paper • 2606.00726 • Published May 30
Uncertainty-Aware Knowledge Distillation for Multimodal Large Language Models Paper • 2603.21426 • Published Mar 22 • 1
Evidence Over Plans: Online Trajectory Verification for Skill Distillation Paper • 2605.09192 • Published May 9 • 1
Mitigating Forgetting Between Supervised and Reinforcement Learning Yields Stronger Reasoners Paper • 2510.04454 • Published Oct 6, 2025
M3-Bench: Multi-Modal, Multi-Hop, Multi-Threaded Tool-Using MLLM Agent Benchmark Paper • 2511.17729 • Published Nov 21, 2025 • 17