Yu Wang
Professor
Works on intelligent chips and energy-efficient circuits and systems, including accelerators and efficient inference for deep learning and large models.
Yu Wang's lab at Tsinghua works on accelerators, chips & architecture, efficient training & inference and large language models. 143 papers since 2023. 20 PhD students and 22 master's students. 66 alumni, most often to ByteDance and Alibaba. Applications close .
At a glance
- papers, 2023–2026
- 37/ year
- at T1 venues
- 0%
- PhD students
- 20
- MS students
- 22
- alumni listed
- 66
Papers by yearfrom the lab homepage, 2023–2026
- 33 papers2026 · T1 0 · collecting
- 49 papers2025 · T1 0
- 30 papers2024 · T1 0
- 31 papers2023 · T1 0
T1 venuesOther
Research focuspaper titles that use each topic's words
- Large language models53 papers
- Efficient training & inference41 papers
- Accelerators, chips & architecture16 papers
- energy-efficient0 papers
- chips0 papers
Paper list
From the lab homepage.
202633 papers
- [CASE26]AED: Automatic Discovery of Effective and Diverse Vulnerabilities for Autonomous Driving Policy with Large Language Models
- [CoRL 26]Beyond Imitation: Reinforcement Learning-Based Sim-Real Co-Training for VLA Models
- [ISCA 26]Bringing Near Data Processing into the Low-Bit Floating-Point Era
- [ICLR26]Cache-to-Cache: Direct Semantic Communication Between Large Language Models
- [ICRA 26]COMPASS: Confined-space Manipulation Planning with Active Sensing Strategy
- [IROS26]D3P: Dynamic Denoising Diffusion Policy via Reinforcement Learning
- [MLSys26]db-SP: Accelerating Sparse Attention for Visual Generative Models with Dual-Balanced Sequence Parallelism
- [OSDI26]DynaRL: Flexible and Dynamic Scheduling of Large-Scale Reinforcement Learning Training
- [EuroSys 26]Efficient and Adaptable Overlapping for Computation and Communication via Signaling and Reordering
- [DATE 26]Endor: Exploit Nearly-Decode-Only Opportunities of LLM Reasoning on Near-Memory Architecture
- [EMNLP 26]HAPMoE: Heterogeneity-Aware Automatic Parallelism Planning for Mixture-of-Experts Models Training
- [RLC26]ICPL: Few-shot In-Context Preference Learning via LLMs
202549 papers
- [TRETS25]CD-LLM: A Heterogeneous Multi-FPGA System for Batched Decoding of 70B+ LLMs using a Compute-Dedicated Architecture
- [CIKM 25]CityLight: A Universal Model Towards Real-world City-scale Traffic Signal Control Coordination
- [ACMMM ]DHCP: Detecting Hallucinations by Cross-modal Attention Pattern in Large Vision-Language Models
- [ICLR 25]Distilled Decoding 1: One-step Sampling of Image Auto-regressive Models with Flow Matching
- [NeurIPS 25]Distilled Decoding 2: One-step Sampling of Image Auto-regressive Models with Conditional Score Distillation
- [IEEE RA-L]Efficient and Hardware-Friendly Online Adaptation for Deep Stereo Depth Estimation on Embedded Robots
- [AAAI ]Enhancing Contrastive Learning Inspired by the Philosophy of ‘the Blind Men and the Elephant’
- [ACMMM ]Fighting Fire with Fire (F3): A Training-free and Efficient Visual Adversarial Example Purification Method in LVLMs
- [IEEE TC]FlashDecoding++Next: High Throughput LLM Inference with Latency and Memory Optimization
- [IEEE RA-L]FlightBench: A Comprehensive Benchmark of Spatial Planning Methods for Quadrotors
- [FPGA25]FlightVGM: Efficient Video Generation Model Inference with Online Sparsification and Hybrid Precision on FPGAs
- [ICCV25]FrameFusion: Combining Similarity and Importance for Video Token Reduction on Large Visual Language Models
202430 papers
- [IEEE TCAD]A Point Transformer Accelerator With Distribution-Aware Heuristic Distance Calculation
- [ICLR24]A Unified Sampling Framework for Solver Searching of Diffusion Probabilistic Models
- [AAAI24]Accelerate Multi-Agent Reinforcement Learning in Zero-Sum Games with Subgame Curriculum Learning
- [DAC24]Automatic Hardware/Software Design for High-Speed Autonomous Unmanned Aerial Vehicles Guided by a Flight Model
- [NeurIPS24]Can LLMs Learn by Teaching? A Preliminary Study
- [IROS24 ]CATOA: Cooperative Calibration of Timestamp Measurements for Distributed Multi-Robot Localization
- [WSDM24 Oral]DeSCo: Towards Generalizable and Scalable Deep Subgraph Counting
- [NeurIPS24]DiTFastAttn: Attention Compression for Diffusion Transformer Models
- [DATE24]DyPIM: Dynamic-inference-enabled Processing-In-Memory Accelerator
- [DAC24]DySpMM: From Fix to Dynamic for Sparse Matrix-Matrix Multiplication Accelerators
- [SOCC24 Special Session]Efficient Deployment of Large Language Model Across Cloud-Device Systems
- [ICML24]Evaluating Quantized Large Language Models
202331 papers
- [IEEE TPAMI]A Generic Graph-based Neural Architecture Encoding Scheme with Multifaceted Information
- [ICCAD23]A Point Transformer Accelerator with Fine-Grained Pipelines and Distribution-Aware Dynamic FPS
- [IEEE RA-L]Active Neural Topological Mapping for Multi-Agent Exploration
- [ICCV23]Ada3D : Exploiting the Spatial Redundancy with Adaptive Inference for Efficient 3D Object Detection
- [DAC23]An Efficient Accelerator for Point-based and Voxel-based Point Cloud Neural Networks
- [AAMAS23]Asynchronous Multi-Agent Reinforcement Learning for Efficient Real-Time Multi-Robot Cooperative Exploration
- [DATE23]CLAP: Locality Aware and Parallel Triangle Counting with Content Addressable Memory
- [IEEE TCAD]CoGNN: An Algorithm-Hardware Co-Design Approach to Accelerate GNN Inference with Mini-Batch Sampling
- [MICRO23]DF-GAS: a Distributed FPGA-as-a-Service Architecture towards Billion-Scale Graph-based Approximate Nearest Neighbor Search
- [AAAI23]Dynamic Ensemble of Low-fidelity Experts: Mitigating NAS
- [AAAI23]Ensemble-in-One: Ensemble Learning within Random Gated Networks for Enhanced Adversarial Robustness
- [MLSys23]Exploiting Hardware Utilization and Adaptive Dataflow for Sparse Convolution in 3D Point Clouds
Where alumni went66 listed, 84% to industry
Industry 32Academia 6
Most often to
- ByteDance
- Alibaba
- Huawei
Also to
- Carnegie Mellon University
- HKUST
- Shanghai Artificial Intelligence Laboratory
Apply to TsinghuaFall 2027
Computer Science PhD (international applicants)
- Round
- Main
- Opens
- GRE
- See program page
- English
- Per school catalog
- Fee
- RMB 800
Not yet confirmed for this intake.
Similar labsnearest by research summary
- USUT Austin
Neeraja Yadwadkar
- Efficient training & inference
- USNYU
Jinyang Li
- Distributed, cloud & edge systems
- USUCLA
Harry Xu
- Distributed, cloud & edge systems
- USUSC
Viktor K. Prasanna
- Accelerators, chips & architecture
- USUW
Stephanie Wang
- Efficient training & inference
- CNSJTU
Chao Li
- Accelerators, chips & architecture
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The lab site and the paper record do not agree.
- This site is shared with other labs, so counts may overlap.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
This lab shares a website with other labs. Member and alumni counts may include them.
Tiers are venue classifications, not lab rankings.
Tsinghuacloses