Meng Li
助理教授
Researches efficient and secure multimodal AI acceleration algorithms and systems.
Meng Li's lab at Peking works on efficient training & inference, vision-language & multimodal and multimodal & grounded language. 67 papers since 2023, 7 at top venues, mostly MICRO and CCS. 17 PhD students and 9 master's students. 8 alumni. Applications close .
At a glance
- papers, 2023–2026
- 17/ year
- at T1 venues
- 10%
- PhD students
- 17
- MS students
- 9
- alumni listed
- 8
Papers by yearfrom the lab homepage, 2023–2026
- 24 papers2026 · T1 5 · collecting
- 20 papers2025 · T1 2
- 14 papers2024 · T1 0
- 9 papers2023 · T1 0
T1 venuesOtherVenues: MICRO, CCS, CoRL, S&P, USENIX Security
Research focuspaper titles that use each topic's words
- Efficient training & inference44 papers
- Vision-language & multimodal8 papers
- acceleration8 papers
- Multimodal & grounded language5 papers
- secure1 paper
Paper list
From the lab homepage.
202624 papers
- International Symposium of Electronic Design Automation (ISEDA)Aging Aware Adaptive Voltage Scaling for Reliable and Efficient AI Accelerators (opens in a new tab)
- International Conference on Machine Learning (ICML)Attention Sink Forges Native MoE in Attention Layers: Sink-Aware Training to Address Head Collapse (opens in a new tab)
- USENIX Symposium on Operating Systems Design and Implementation (OSDI)Breaking the Reward Barrier: Accelerating Tree-of-Thought Reasoning via Speculative Exploration (opens in a new tab)
- S&P (T1)Cachemir: Fully Homomorphic Encrypted Inference of Generative Large Language Model with KV Cache (opens in a new tab)
- International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS)CREATE: Cross-Layer Resilience Characterization and Optimization for Efficient yet Reliable Embodied AI Systems (opens in a new tab)
- Design Automation Conference (DAC)DRIFT: Harnessing Inherent Fault Tolerance for Efficient and Reliable Diffusion Model Inference (opens in a new tab)
- Design Automation Conference (DAC)DySL-VLA: Efficient Vision-Language-Action Model Inference via Dynamic-Static Layer-Skipping for Robot Manipulation (opens in a new tab)
- Design Automation Conference (DAC)EdgeSC: Universal Stochastic Computing Architecture for Efficient Edge Detection (opens in a new tab)
- IEEE TCADHDA-MoE: Hybrid Parallelism and Dynamic, Adaptive Scheduling for Mixture-of-Experts with 3D Near-Memory Processing (opens in a new tab)
- MICRO (T1)Helios: Melting Kernel Boundaries for GPU-Accelerated HE via Graph Rewriting and Microarchitecture-Aware Mapping (opens in a new tab)
- International Conference on Machine Learning (ICML)HyPER: Bridging Exploration and Exploitation for Scalable LLM Reasoning with Hypothesis Path Expansion and Reduction (opens in a new tab)
- CoRL (T1)Jetson-PI: Towards Onboard Real-Time Robot Control via Foresight-Aligned Asynchronous Inference (opens in a new tab)
202520 papers
- Asian Solid-State Circuits Conference (A-SSCC)A 28nm 534.6TOPS/W Mixed-Precision Edge Accelerator for Embodied AI Using Stochastic Computing (opens in a new tab)
- USENIX Security (T1)Breaking the Layer Barrier: Remodeling Private Transformer Inference with Hybrid CKKS and MPC (opens in a new tab)
- Design, Automation and Test in Europe Conference and Exhibition (DATE)Compact Non-Volatile Lookup Table Architecture based on Ferroelectric FET Array through In-Situ Combinatorial One-Hot Encoding for Reconfigurable Computing (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPs)CryptoMoE: Privacy-Preserving and Scalable Mixture of Experts Inference via Balanced Expert Routing (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPs)EfficientNav: Towards On-Device Object-Goal Navigation with Navigation Map Caching and Retrieval (opens in a new tab)
- International Conference on Computer-Aided Design (ICCAD)FENIX: Flexible and Efficient Hybrid HE/MPC Acceleration with Near-Memory Processing (opens in a new tab)
- Design, Automation and Test in Europe Conference and Exhibition (DATE)FLASH: An Efficient Hardware Accelerator Leveraging Approximate and Sparse FFT for Homomorphic Encryption (opens in a new tab)
- International Conference on Computer-Aided Design (ICCAD)H2EAL: Hybrid-Bonding Architecture with Hybrid Sparse Attention for Efficient Long-Context LLM Inference (opens in a new tab)
- International Conference on Computer-Aided Design (ICCAD)HD-MoE: Hybrid and Dynamic Parallelism for Mixture-of-Expert LLMs with 3D Near-Memory Processing (opens in a new tab)
- Design Automation Conference (DAC)HybriMoE: Hybrid CPU-GPU Scheduling and Cache Management for Efficient MoE Inference (opens in a new tab)
- MICRO (T1)Ironman: Accelerating Oblivious Transfer Extension for Privacy-Preserving AI with Near-Memory Processing (opens in a new tab)
- Design, Automation and Test in Europe Conference and Exhibition (DATE)LightMamba: Efficient Mamba Acceleration on FPGA with Quantization and Hardware Co-design (opens in a new tab)
202414 papers
- IEEE Transactions on Circuits and Systems I: Regular Papers (TCAS-I)A 16.38TOPS and 4.55POPS/W SRAM Computing-in-Memory Macro for Signed Operands Computation and Batch Normalization Implementation (opens in a new tab)
- International Conference on Computer-Aided Design (ICCAD)AdapMoE: Adaptive Sensitivity-based Expert Gating and Management for Efficient MoE Inference (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPs)ArkVale: Efficient Generative LLM Inference with Recallable Key-Value Eviction (opens in a new tab)
- Design, Automation and Test in Europe Conference and Exhibition (DATE)ASCEND: Accurate yet Efficient End-to-End Stochastic Computing Acceleration of Vision Transformer (opens in a new tab)
- IEEE Transactions on Circuits and Systems I: Regular Papers (TCAS-I)CASCADE: A Framework for CNN Accelerator Synthesis with Concatenation and Refreshing Dataflow (opens in a new tab)
- Design Automation Conference (DAC)FastQuery: Communication-efficient Embedding Table Query for Private LLMs inference (opens in a new tab)
- International Conference on Computer-Aided Design (ICCAD)FlexHE: A flexible Kernel Generation Framework for Homomorphic Encryption-Based Private Inference (opens in a new tab)
- International Conference on Computer-Aided Design (ICCAD)HG-PIPE: Vision Transformer Acceleration with Hybrid-Grained Pipeline (opens in a new tab)
- International Conference on Computer-Aided Design (ICCAD)MCUBERT: Memory-Efficient BERT Inference on Commodity Microcontrollers (opens in a new tab)
- IEEE Custom Integrated Circuits Conference (CICC)MixCIM: A Hybrid-Cell-Based Computing-in-Memory Macro with Less-Data-Movement and Activation-Memory-Reuse for Depthwise Separable Neural Networks (opens in a new tab)
- International Conference on Computer-Aided Design (ICCAD)OSCA: End-to-end Serial Stochastic Computing Neural Acceleration with Fine-grained Scaling and Piecewise Activation (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPs)PrivCirNet: Efficient Private Inference via Block Circulant Transformation (opens in a new tab)
20239 papers
- Design, Automation and Test in Europe Conference and Exhibition (DATE)Accurate yet Efficient Stochastic Computing Neural Acceleration with High Precision Residual Fusion (opens in a new tab)
- IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD)AVATAR: An Aging- and Variation-Aware Dynamic Timing Analyzer for Error-Efficient Computing (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPs)CoPriv: Network/Protocol Co-Optimization for Communication-Efficient Private Inference (opens in a new tab)
- Design Automation Conference (DAC)Efficient Non-Linear Adder for Stochastic Computing with Approximate Spatial-Temporal Sorting Network (opens in a new tab)
- ACM/IEEE International Conference on Computer Aided Design (ICCAD)Falcon: Accelerating Homomorphically Encrypted Convolutions for Efficient Private Mobile Network Inference (opens in a new tab)
- ACM/IEEE International Conference on Computer Aided Design (ICCAD)Memory-aware Scheduling for Complex Wired Networks with Iterative Graph Optimization (opens in a new tab)
- International Conference on Computer Vision (ICCV)MPCViT: Searching for Accurate and Efficient MPC-Friendly Vision Transformer with Heterogeneous Attention (opens in a new tab)
- International Conference on ASIC (ASICON)Not your father’s stochastic computing (SC)! Efficient yet Accurate End-to-End SC Accelerator Design (opens in a new tab)
- ACM/IEEE International Conference on Computer Aided Design (ICCAD)READ: Reliability-Enhanced Accelerator Dataflow Optimization using Critical Input Pattern Reduction (opens in a new tab)
Where alumni went8 listed
Apply to PekingFall 2027
Computer Science PhD (international applicants)
- 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
- USUMD
Jennifer Golbeck
- Computational social science
- CNPeking
Leye Wang
- Time series, anomaly & spatio-temporal data
- USCMU
Nikolas Martelaro
- Interfaces & interaction techniques
- USGeorgia Tech
Haesun Park
- Visualization & visual analytics
- CNZhejiang
Wei Chen
- Medical image analysis
- CHETH Zurich
Ulrik Brandes
- Graph mining & networks
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
Pekingcloses