8 papers
Deep Ensembling with No Overhead for either Training or Testing: The All-Round Blessings of Dynamic Sparsity
Shiwei Liu, Tianlong Chen, Zahra Atashgahi +6
The success of deep ensembles on improving predictive performance, uncertainty estimation, and out-of-distribution robustness has been extensively studied in the machine learning l…
ResearchGPT: Benchmarking and Training LLMs for End-to-End Computer Science Research Workflows
Penghao Wang, Yuhao Zhou, Mengxuan Wu +12
As large language models (LLMs) advance, the ultimate vision for their role in science is emerging: we could build an AI collaborator to effectively assist human beings throughout…
One Token Embedding Is Enough to Deadlock Your Large Reasoning Model
Mohan Zhang, Yihua Zhang, Jinghan Jia +3
Modern large reasoning models (LRMs) exhibit impressive multi-step problem-solving via chain-of-thought (CoT) reasoning. However, this iterative thinking mechanism introduces a new…
More is Less: The Pitfalls of Multi-Model Synthetic Preference Data in DPO Safety Alignment
Yifan Wang, Runjin Chen, Bolian Li +7
Aligning large language models (LLMs) with human values is an increasingly critical step in post-training. Direct Preference Optimization (DPO) has emerged as a simple, yet effecti…
Finding Fantastic Experts in MoEs: A Unified Study for Expert Dropping Strategies and Observations
Ajay Jaiswal, Jianyu Wang, Yixiao Li +6
Sparsely activated Mixture-of-Experts (SMoE) has shown promise in scaling up the learning capacity of neural networks. However, vanilla SMoEs have issues such as expert redundancy…
DANCE: DAta-Network Co-optimization for Efficient Segmentation Model Training and Inference
Chaojian Li, Wuyang Chen, Yuchen Gu +4
Semantic segmentation for scene understanding is nowadays widely demanded, raising significant challenges for the algorithm efficiency, especially its applications on resource-limi…