Yang You
NUS Presidential Young Professor
Leads the HPC-AI Lab on scaling deep neural network training on distributed systems, including the LAMB optimiser.
Yang You's lab at NUS works on efficient training & inference and distributed, cloud & edge systems. 85 papers since 2023, 64 at top venues, mostly NeurIPS and ICLR. 13 PhD students and 2 master's students. 29 alumni, most often to Google DeepMind and HPC-AI Technology. Applications close .
At a glance
- papers, 2023–2026
- 22/ year
- at T1 venues
- 75%
- PhD students
- 13
- MS students
- 2
- alumni listed
- 29
Papers by yearfrom the lab homepage and DBLP, 2023–2026
- 12 papers2026 · T1 10 · collecting
- 24 papers2025 · T1 21
- 24 papers2024 · T1 17
- 25 papers2023 · T1 16
T1 venuesOtherVenues: NeurIPS, ICLR, ICML, CVPR, AAAI
Research focuspaper titles that use each topic's words
- Efficient training & inference32 papers
- Distributed, cloud & edge systems4 papers
- scaling2 papers
- network training0 papers
- deep neural0 papers
Paper list
From the lab homepage and DBLP.
202612 papers
- ICLR (T1)ASCIIEval: Benchmarking Models' Visual Perception in Text Strings via ASCII Art (opens in a new tab)
- ACM Trans. Archit. Code Optim.CLAP: Cross-Layer Adaptive Pipelining Inference Scheduling for Resource-Efficient Edge-Cloud Vision Systems (opens in a new tab)
- TPAMI (T1)DyDiT++: Diffusion Transformers With Timestep and Spatial Dynamics for Efficient Visual Generation (opens in a new tab)
- CVPR (T1)Efficient Video Object Segmentation and Tracking with Recurrent Dynamic Submodel (opens in a new tab)
- ICLR (T1)FOCUS: Efficient Keyframe Selection for Long Video Understanding (opens in a new tab)
- ACL (T1)GroupToM-Bench: Benchmarking Group Theory of Mind and Nonlinear Social Emergence in MLLMs (opens in a new tab)
- PPoPPHelixPipe: Efficient Distributed Training of Long Sequence Transformers with Attention Parallel Pipeline Parallelism (opens in a new tab)
- ICLR (T1)K-Sort Eval: Efficient Preference Evaluation for Visual Generation via Corrected VLM-as-a-Judge (opens in a new tab)
- AAAI (T1)MDK12-Bench: A Multi-Discipline Benchmark for Evaluating Reasoning in Multimodal Large Language Models (opens in a new tab)
- ICLR (T1)MoNE: Replacing Redundant Experts with Lightweight Novices for Structured Pruning of MoE (opens in a new tab)
- ICLR (T1)RAPID3: Tri-Level Reinforced Acceleration Policies for Diffusion Transformer (opens in a new tab)
- CVPR (T1)Region-Adaptive Sampling for Diffusion Transformers (opens in a new tab)
202524 papers
- CVPR (T1)A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training (opens in a new tab)
- CVPR (T1)A Stitch in Time Saves Nine: Small VLM is a Precise Guidance for Accelerating Large VLMs (opens in a new tab)
- ASPLOS (T1)Concerto: Automatic Communication Optimization and Scheduling for Large-Scale Deep Learning (opens in a new tab)
- ACL (T1)DavIR: Data Selection via Implicit Reward for Large Language Models (opens in a new tab)
- NeurIPS (T1)Drag-and-Drop LLMs: Zero-Shot Prompt-to-Weights (opens in a new tab)
- ICML (T1)DSP: Dynamic Sequence Parallelism for Multi-Dimensional Transformers (opens in a new tab)
- ICLR (T1)Dynamic Diffusion Transformer (opens in a new tab)
- NeurIPS (T1)ElasticMM: Efficient Multimodal LLMs Serving with Elastic Multimodal Parallelism (opens in a new tab)
- CVPR (T1)Emphasizing Discriminative Features for Dataset Distillation in Complex Scenarios (opens in a new tab)
- COLMEnsemble Debiasing Across Class and Sample Levels for Fairer Prompting Accuracy
- ICML (T1)Info-Coevolution: An Efficient Framework for Data Model Coevolution (opens in a new tab)
- IEEE Trans. Neural Networks Learn. Syst.KMT-PLL: K-Means Cross-Attention Transformer for Partial Label Learning (opens in a new tab)
202424 papers
- ICLR (T1)AutoChunk: Automated Activation Chunk for Memory-Efficient Deep Learning Inference (opens in a new tab)
- ECCV (T1)Dataset Growth (opens in a new tab)
- ICML (T1)DiffAug: Enhance Unsupervised Contrastive Learning with Domain-Knowledge-Free Diffusion-based Data Augmentation (opens in a new tab)
- ECCV WorkshopDiM: Distilling Dataset into Generative Model (opens in a new tab)
- TKDE (T1)Distributed and Joint Evidential K-Nearest Neighbor Classification (opens in a new tab)
- NeurIPS (T1)Dynamic Tuning Towards Parameter and Inference Efficiency for ViT Adaptation (opens in a new tab)
- CVPR (T1)Efficient Dataset Distillation via Minimax Diffusion (opens in a new tab)
- PPoPPFastFold: Optimizing AlphaFold Training and Inference on GPU Clusters (opens in a new tab)
- WWW (T1)Helen: Optimizing CTR Prediction Models with Frequency-wise Hessian Eigenvalue Regularization (opens in a new tab)
- MLSys (T1)HeteGen: Efficient Heterogeneous Parallel Inference for Large Language Models on Resource-Constrained Devices (opens in a new tab)
- EMNLP (T1)How Does the Textual Information Affect the Retrieval of Multimodal In-Context Learning? (opens in a new tab)
- ICLR (T1)InfoBatch: Lossless Training Speed Up by Unbiased Dynamic Data Pruning (opens in a new tab)
202325 papers
- Knowl. Based Syst.A new multi-source Transfer Learning method based on Two-stage Weighted Fusion (opens in a new tab)
- TKDE (T1)A Sparse Reconstructive Evidential K-Nearest Neighbor Classifier for High-Dimensional Data (opens in a new tab)
- ICML (T1)A Study on Transformer Configuration and Training Objective (opens in a new tab)
- ICML (T1)Adaptive Computation with Elastic Input Sequence (opens in a new tab)
- Information SciencesAdaptive evidential K-NN classification: Integrating neighborhood search and feature weighting (opens in a new tab)
- IPDPSAn Efficient 2D Method for Training Super-Large Deep Learning Models (opens in a new tab)
- CVPR (T1)BiCro: Noisy Correspondence Rectification for Multi-modality Data via Bi-directional Cross-modal Similarity Consistency (opens in a new tab)
- ACL (T1)CAME: Confidence-guided Adaptive Memory Efficient Optimization (opens in a new tab)
- ICPPColossal-AI: A Unified Deep Learning System For Large-Scale Parallel Training (opens in a new tab)
- AAAI (T1)CowClip: Reducing CTR Prediction Model Training Time from 12 Hours to 10 Minutes on 1 GPU (opens in a new tab)
- ICCV (T1)Dataset Quantization (opens in a new tab)
- ICLR (T1)Divide to Adapt: Mitigating Confirmation Bias for Domain Adaptation of Black-Box Predictors (opens in a new tab)
Where alumni went29 listed, 100% to industry
Industry 3Academia 0
Most often to
- Google DeepMind
- HPC-AI Technology
- Tencent
Apply to NUSFall 2027
School of Computing PhD in Computer Science
- Round
- Main
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CNZhejiang
Kun Yang
- Accelerators, chips & architecture
- DETUM
Markus Götz
- Distributed, cloud & edge systems
- USUSC
Murali Annavaram
- Efficient training & inference
- USUT Austin
Diana Marculescu
- Accelerators, chips & architecture
- USUSC
Priyadarshini Panda
- Accelerators, chips & architecture
- USUSC
Seo Jin Park
- Efficient training & inference
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage and DBLP
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
NUScloses