Danny Liu
Also known as Guan-Ting Liu
Danny Liu is a researcher at NVIDIA Research Taiwan whose work connects reinforcement learning, chip design automation, and language models for hardware. He earned a bachelor's degree in electronic engineering with a computer-science cross-discipline at NCTU and completed a PhD in networking and multimedia at National Taiwan University. His papers traced a path from cross-state and programmatic reinforcement learning to neural models for chip timing and reasoning language models for RTL code. As of October 2026, he is a member of NVIDIA Research Taiwan, a group focused on generative AI, multimodal learning, and 3D vision.
Timeline
-
By 2022
Cross-state behavior matching
Before graduate study at National Taiwan University, he earns a bachelor's degree in Electronic Engineering with a cross-discipline in Computer Science at NCTU. In 2022, he coauthors the AAMAS extended abstract “Improving Generalization with Cross-State Behavior Matching in Deep Reinforcement Learning.”
-
2023
Programmatic policies for RL
At ICML 2023, he and his coauthors introduce hierarchical programmatic reinforcement learning, which composes programs from a learned embedding space into policies for complex behaviors outside the training distribution.
-
2024
A PhD and timing models
In 2024, he completes a PhD in National Taiwan University's Graduate Institute of Networking and Multimedia, advised by Pu-Jen Cheng, with a dissertation on generalizable and interpretable reinforcement learning. He also coauthors a neural collaborative filtering study of multi-corner timing macro modeling.
-
2025
ScaleRTL for RTL coding
In 2025, he coauthors ScaleRTL, a reasoning language model for RTL coding that scales high-quality reasoning data and test-time compute.
-
2026
ACE-RTL for hardware design
In 2026, he coauthors ACE-RTL, which combines an RTL-specialized language model with a reasoning model and iteratively refines hardware code.
-
As of October 2026
NVIDIA Research Taiwan
As of October 2026, he is a member of NVIDIA Research Taiwan. His research interests include reinforcement learning, machine learning for EDA, and generative AI; the team focuses on generative AI, multimodal learning, and 3D vision.
Career
- Since 2025NVIDIAResearcher
Education
- 2024National Taiwan UniversityPhD
- National Chiao Tung UniversityBachelor's degree
Most-cited works
- Construction of Fe0.64Ni0.36@graphite nanoparticles via corrosion-like transformation from NiFe2O4 and surface graphitization in flexible carbon nanofibers to achieve strong wideband microwave absorption
- Inner-Bond-Cleavage Approach to Figure-Eight Macrocycles from Planar Aromatic Hydrocarbons
- Flexible electrospun iron/manganese-based compounds/carbon fibers: Phase transformation and electrochemical properties
- Low Equivalent Oxide Thickness and Leakage Current of pGe MOS Device by Removing Low Oxidation State in GeOx With H2 Plasma Treatment
- Thermally activated delayed fluorescence carbazole‐triazine dendrimer with bulky substituents
Colleagues
Sources and credits
Photo: We have no published photo of Danny Liu yet. Send us a free-licensed photo.
- (Design Automation Research) Danny Liu | NVIDIA Research Taiwan
- AAMAS 2022: Proceedings of the 2022 International Conference on Autonomous Agents and Multiagent Systems
- Hierarchical Programmatic Reinforcement Learning via Learning to Compose Programs
- NTU Theses and Dissertations Repository: 朝向可泛化和可解釋的強化學習
- Multi-Corner Timing Macro Modeling With Neural Collaborative Filtering From Recommendation Systems Perspective
- [2506.05566] ScaleRTL: Scaling LLMs with Reasoning Data and Test-Time Compute for Accurate RTL Code Generation
- [2602.10218] ACE-RTL: When Agentic Context Evolution Meets RTL-Specialized LLMs
- NVIDIA Research Taiwan