David Z. Pan
YOU?
Author Swipe
View article: ENLighten: Lighten the Transformer, Enable Efficient Optical Acceleration
ENLighten: Lighten the Transformer, Enable Efficient Optical Acceleration Open
Photonic computing has emerged as a promising substrate for accelerating the dense linear-algebra operations at the heart of AI, yet adoption for large Transformer models remains in its infancy. We identify two bottlenecks: (1) costly elec…
View article: Fine-Tuning Masked Diffusion for Provable Self-Correction
Fine-Tuning Masked Diffusion for Provable Self-Correction Open
A natural desideratum for generative models is self-correction--detecting and revising low-quality tokens at inference. While Masked Diffusion Models (MDMs) have emerged as a promising approach for generative modeling in discrete spaces, t…
View article: TopoSizing: An LLM-aided Framework of Topology-based Understanding and Sizing for AMS Circuits
TopoSizing: An LLM-aided Framework of Topology-based Understanding and Sizing for AMS Circuits Open
Analog and mixed-signal circuit design remains challenging due to the shortage of high-quality data and the difficulty of embedding domain knowledge into automated flows. Traditional black-box optimization achieves sampling efficiency but …
View article: AutoMarks: A GNN-based Automated Physical Design Watermarking Framework
AutoMarks: A GNN-based Automated Physical Design Watermarking Framework Open
Physical design refers to converting a logical circuit description into a physical chip layout for manufacturing. Although design companies invest huge efforts to fine-tune the cell positions and connections for better power, performance, …
View article: Hardware-efficient photonic tensor core: accelerating deep neural networks with structured compression
Hardware-efficient photonic tensor core: accelerating deep neural networks with structured compression Open
The rapid growth in computing demands, particularly driven by artificial intelligence applications, has begun to exceed the capabilities of traditional electronic hardware. Optical computing offers a promising alternative due to its parall…
View article: Hardware-efficient photonic tensor core: accelerating deep neural networks with structured compression
Hardware-efficient photonic tensor core: accelerating deep neural networks with structured compression Open
The rapid growth in computing demands, particularly driven by artificial intelligence applications, has begun to exceed the capabilities of traditional electronic hardware. Optical computing offers a promising alternative due to its parall…
View article: UniMoCo: Unified Modality Completion for Robust Multi-Modal Embeddings
UniMoCo: Unified Modality Completion for Robust Multi-Modal Embeddings Open
Current research has explored vision-language models for multi-modal embedding tasks, such as information retrieval, visual grounding, and classification. However, real-world scenarios often involve diverse modality combinations between qu…
View article: AnalogCoder: Analog Circuit Design via Training-Free Code Generation
AnalogCoder: Analog Circuit Design via Training-Free Code Generation Open
Analog circuit design is a significant task in modern chip technology, focusing on the selection of component types, connectivity, and parameters to ensure proper circuit functionality. Despite advances made by Large Language Models (LLMs)…
View article: Can Test-Time Scaling Improve World Foundation Model?
Can Test-Time Scaling Improve World Foundation Model? Open
World foundation models, which simulate the physical world by predicting future states from current observations and inputs, have become central to many applications in physical intelligence, including autonomous driving and robotics. Howe…
View article: Late Breaking Results: Breaking Symmetry- Unconventional Placement of Analog Circuits using Multi-Level Multi-Agent Reinforcement Learning
Late Breaking Results: Breaking Symmetry- Unconventional Placement of Analog Circuits using Multi-Level Multi-Agent Reinforcement Learning Open
Layout-dependent effects (LDEs) significantly impact analog circuit performance. Traditionally, designers have relied on symmetric placement of circuit components to mitigate variations caused by LDEs. However, due to non-linear nature of …
View article: DICE: Device-level Integrated Circuits Encoder with Graph Contrastive Pretraining
DICE: Device-level Integrated Circuits Encoder with Graph Contrastive Pretraining Open
Pretraining models with unsupervised graph representation learning has led to significant advancements in domains such as social network analysis, molecular design, and electronic design automation (EDA). However, prior work in EDA has mai…
View article: Vchitect-2.0: Parallel Transformer for Scaling Up Video Diffusion Models
Vchitect-2.0: Parallel Transformer for Scaling Up Video Diffusion Models Open
We present Vchitect-2.0, a parallel transformer architecture designed to scale up video diffusion models for large-scale text-to-video generation. The overall Vchitect-2.0 system has several key designs. (1) By introducing a novel Multimod…
View article: APOLLO: SGD-like Memory, AdamW-level Performance
APOLLO: SGD-like Memory, AdamW-level Performance Open
Large language models (LLMs) are notoriously memory-intensive during training, particularly with the popular AdamW optimizer. This memory burden necessitates using more or higher-end GPUs or reducing batch sizes, limiting training scalabil…
View article: M3: Mamba-assisted Multi-Circuit Optimization via MBRL with Effective Scheduling
M3: Mamba-assisted Multi-Circuit Optimization via MBRL with Effective Scheduling Open
Recent advancements in reinforcement learning (RL) for analog circuit optimization have demonstrated significant potential for improving sample efficiency and generalization across diverse circuit topologies and target specifications. Howe…
View article: PACE: Pacing Operator Learning to Accurate Optical Field Simulation for Complicated Photonic Devices
PACE: Pacing Operator Learning to Accurate Optical Field Simulation for Complicated Photonic Devices Open
Electromagnetic field simulation is central to designing, optimizing, and validating photonic devices and circuits. However, costly computation associated with numerical simulation poses a significant bottleneck, hindering scalability and …
View article: SELA: Tree-Search Enhanced LLM Agents for Automated Machine Learning
SELA: Tree-Search Enhanced LLM Agents for Automated Machine Learning Open
Automated Machine Learning (AutoML) approaches encompass traditional methods that optimize fixed pipelines for model selection and ensembling, as well as newer LLM-based frameworks that autonomously build pipelines. While LLM-based agents …
View article: Selecting robust silicon photonic designs after Bayesian optimization without extra simulations
Selecting robust silicon photonic designs after Bayesian optimization without extra simulations Open
Optimization methods are frequently exploited in the design of silicon photonic devices. In this paper, we demonstrate that pushing the objective function to its minimum during optimization often results in devices that gradually become mo…
View article: Open-Source Differentiable Lithography Imaging Framework
Open-Source Differentiable Lithography Imaging Framework Open
The rapid evolution of the electronics industry, driven by Moore's law and the proliferation of integrated circuits, has led to significant advancements in modern society, including the Internet, wireless communication, and artificial inte…
View article: Differentiable Edge-based OPC
Differentiable Edge-based OPC Open
Optical proximity correction (OPC) is crucial for pushing the boundaries of semiconductor manufacturing and enabling the continued scaling of integrated circuits. While pixel-based OPC, termed as inverse lithography technology (ILT), has g…
View article: Automated Physical Design Watermarking Leveraging Graph Neural Networks
Automated Physical Design Watermarking Leveraging Graph Neural Networks Open
This paper presents AutoMarks, an automated and transferable watermarking framework that leverages graph neural networks to reduce the watermark search overheads during the placement stage. AutoMarks's novel automated watermark search is a…
View article: INSIGHT: Universal Neural Simulator for Analog Circuits Harnessing Autoregressive Transformers
INSIGHT: Universal Neural Simulator for Analog Circuits Harnessing Autoregressive Transformers Open
Analog front-end design heavily relies on specialized human expertise and costly trial-and-error simulations, which motivated many prior works on analog design automation. However, efficient and effective exploration of the vast and comple…