11 papers
On the Pitfalls of Verbalized Confidence Priors for Calibrating Large Reasoning Models
Shuoyuan Wang, Beier Luo, Hao Zeng +5
Large reasoning models (LRMs) often suffer from overconfidence when expressing their uncertainty. Confidence-aware reinforcement learning (RL) offers a promising way to optimize ca…
Anytime-Valid LLM Leaderboards via Benchmark-weighted and Block-Factorized e-Processes
Hongfu Gao, Songxin Zhang, Zejian Xie +3
Large language model (LLM) leaderboards compare model capabilities by ranking models according to their mean performance on fixed benchmarks. However, variability in evaluation out…
AtomEgo: Exploring Ego-Robot Integration for Embodied Foundation Model Pretraining
Di Wu, Dongchen Zheng, Junhe Sheng +8
Embodied foundation models are constrained by the limited scale and diversity of robot demonstrations, motivating the use of large-scale egocentric human interaction data. However,…
HSAP: A Hierarchical Sequence-aware Parallelism for Hybrid-Context Generative Models
Songxin Zhang, Zejian Xie, Zhuoyang Song +4
In this paper, we aim to combine the advantages of existing sequence parallelism paradigms and overcomes their drawbacks, the most serious of which is the incapability to correctly…
BatchWeave: A Consistent Object-Store-Native Data Plane for Large Foundation Model Training
Ting Sun, Junjie Zhang, Xiao Yan +7
Modern Large Foundation Model (LFM) training has transformed the data pipeline from a static ingestion layer into a dynamic component that must co-evolve with the training process.…
PVI: Plug-in Visual Injection for Vision-Language-Action Models
Zezhou Zhang, Songxin Zhang, Xiao Xiong +8
VLA architectures that pair a pretrained VLM with a flow-matching action expert have emerged as a strong paradigm for language-conditioned manipulation. Yet the VLM, optimized for…
Natural Language-Driven Global Mapping of Martian Landforms
Yiran Wang, Shuoyuan Wang, Zhaoran Wei +7
Planetary surfaces are typically analyzed using high-level semantic concepts in natural language, yet vast orbital image archives remain organized at the pixel level. This mismatch…
Orcust: Stepwise-Feedback Reinforcement Learning for GUI Agent
Junyu Lu, Songxin Zhang, Zejian Xie +2
Recent advances in GUI agents have achieved remarkable grounding and action-prediction performance, yet existing models struggle with unreliable reward signals and limited online t…
L0: Reinforcement Learning to Become General Agents
Junjie Zhang, Jingyi Xi, Zhuoyang Song +7
Training large language models (LLMs) to act as autonomous agents for multi-turn, long-horizon tasks remains significant challenges in scalability and training efficiency. To addre…
Astrea: A MOE-based Visual Understanding Model with Progressive Alignment
Xiaoda Yang, JunYu Lu, Hongshun Qiu +12
Vision-Language Models (VLMs) based on Mixture-of-Experts (MoE) architectures have emerged as a pivotal paradigm in multimodal understanding, offering a powerful framework for inte…
Exploring Learning Complexity for Efficient Downstream Dataset Pruning
Wenyu Jiang, Zhenlong Liu, Zejian Xie +3
The ever-increasing fine-tuning cost of large-scale pre-trained models gives rise to the importance of dataset pruning, which aims to reduce dataset size while maintaining task per…