Sitemap
A list of all the posts and pages found on the site. For you robots out there, there is an XML version available for digesting as well.
Pages
Posts
Stop Training Your LLM Router: Zero-Shot Confidence Beats Supervised Baselines
Published:
TL;DR: We tested whether you actually need labeled training data to route queries between a cheap local LLM and an expensive cloud model. You don’t. Average token log-probability — available for free from the first query — matches supervised routing in-distribution and crushes it when the query distribution shifts. We tested across 3 model families, 2 datasets, ~4,500 queries, and $123 in cloud costs. All code and data are open.
Things I got wrong building a confidence evaluator for local LLMs
Published:
An in-progress lab-notebook post from the Autodidact project.
Building AI agents using LangChain
Published:
LangChain
What Automotive AI Taught Me About Responsible Deployment
Published:
When you ship AI into a vehicle, the consequences of failure aren’t abstract. They’re not a bad user review or a dip in engagement metrics. They’re a driver who takes their eyes off the road because the system said something confidently wrong, or a voice assistant that crashes mid-navigation and leaves someone lost on an unfamiliar highway. Shipping AI where failures have physical consequences changes how you think about deployment — permanently.
AI Agents: Buzzword or Paradigm Shift?
Published:
“AI agent”—a fancy buzzword for AI model integration, or is there more to it? With increasing autonomy and decision-making capabilities, are we just rebranding models, or is this a true paradigm shift?
Best Student Paper Award at HOST 2020
Published:
I’m excited to share that our paper “A Novel Golden-Chip-Free Clustering Technique Using Backscattering Side Channel for Hardware Trojan Detection” won the Best Student Paper Award at the 2020 IEEE International Symposium on Hardware Oriented Security and Trust (HOST).
Upcoming Talk: Cybersecurity Lecture Series at Georgia Tech
Published:
I’m excited to announce that I will be giving a talk as part of the Cybersecurity Lecture Series at Georgia Tech on January 24, 2020.
notes
Path to expertise: On-Device Speech & Language Processing
Published:
Intermediate → Expert: On-Device Speech & Language Processing
How AI Changed the Way I Do Research and Write Research Papers
Published:
My latest paper, “Zero-Shot Confidence Estimation for Small LLMs: When Supervised Baselines Aren’t Worth Training,” is now published. It’s an empirical paper — a step toward getting back into the research world. However, this post isn’t about the paper itself — it’s about how dramatically the research process has changed since my last one.
portfolio
Autodidact — A Local-First AI Agent That Learns
Open-source AI agent with a local brain: thinks locally first, escalates to the cloud when uncertain, and distills what it learns back down so the lesson is permanent.
GenAI Integration for Safety-Critical Automotive AI
Production LLM integration with safety constraints for automotive voice assistant
On-Device Voice AI Pipeline
End-to-end on-device speech inference pipeline with Conformer ASR and WFST decoding, engineered for strict latency and memory budgets on embedded automotive hardware.
Sparse Autoencoder — Interpretability Reference Implementation
A study-and-build reimplementation of sparse autoencoders for neural network interpretability, following Anthropic’s mechanistic interpretability research.
publications
Advanced Decision of PU Partition and CU Depth for Fractional Motion Estimation in HEVC
Published in International Conference on Electronics, Information and Communication (ICEIC), 2016
This paper presents an advanced early termination algorithm for prediction unit (PU) partition and coding unit (CU) depth decisions in HEVC fractional motion estimation, reducing encoding complexity while preserving rate-distortion performance.
Recommended citation: Nguyen Ngoc Luong, Tae Sung Kim, Hyuk-Jae Lee, Soo-Ik Chae, "Advanced Decision of PU Partition and CU Depth for Fractional Motion Estimation in HEVC," International Conference on Electronics, Information and Communication (ICEIC), Jan. 2016.
A Reduction of Interpolation Redundancy for Fractional Motion Estimation in HEVC
Published in 2016 SoC Conference of Korea, 2016
🏆 Best Paper Award
Recommended citation: Luong N. Nguyen, Tae Sung Kim, Hyuk-Jae Lee, "A Reduction of Interpolation Redundancy for Fractional Motion Estimation in HEVC," 2016 SoC Conference of Korea, May 2016.
Creating a backscattering side channel to enable detection of dormant hardware trojans
Published in IEEE transactions on very large scale integration (VLSI) systems, 2019
Use Google Scholar for full citation
Recommended citation: Luong Nguyen, Chia-Lin Cheng, Milos Prvulovic, Alenka Zaji{\'c}, "Creating a backscattering side channel to enable detection of dormant hardware trojans." IEEE transactions on very large scale integration (VLSI) systems, 2019.
Exploiting switching of transistors in digital electronics for RFID tag design
Published in IEEE Journal of Radio Frequency Identification, 2019
🏆 Best Poster Award (12th International Conference on RFID 2018)
Recommended citation: Chia-Lin Cheng, Luong Nguyen, Milos Prvulovic, Alenka Zaji{\'c}, "Exploiting switching of transistors in digital electronics for RFID tag design." IEEE Journal of Radio Frequency Identification, 2019.
IDEA: Intrusion detection through electromagnetic-signal analysis for critical embedded and cyber-physical systems
Published in IEEE Transactions on Dependable and Secure Computing, 2019
Use Google Scholar for full citation
Recommended citation: Haider Khan, Nader Sehatbakhsh, Luong Nguyen, Robert Callan, Arie Yeredor, Milos Prvulovic, Alenka Zaji{\'c}, "IDEA: Intrusion detection through electromagnetic-signal analysis for critical embedded and cyber-physical systems." IEEE Transactions on Dependable and Secure Computing, 2019.
Malware detection in embedded systems using neural network model for electromagnetic side-channel signals
Published in Journal of Hardware and Systems Security, 2019
Use Google Scholar for full citation
Recommended citation: Haider Khan, Nader Sehatbakhsh, Luong Nguyen, Milos Prvulovic, Alenka Zaji{\'c}, "Malware detection in embedded systems using neural network model for electromagnetic side-channel signals." Journal of Hardware and Systems Security, 2019.
A comparison of backscattering, EM, and power side-channels and their performance in detecting software and hardware intrusions
Published in Journal of Hardware and Systems Security, 2020
Use Google Scholar for full citation
Recommended citation: L. Nguyen, C. Cheng, F. Werner, M. Prvulovic, and A. Zajic, "A comparison of backscattering, EM, and power side-channels and their performance in detecting software and hardware intrusions," Journal of Hardware and Systems Security, March 2020.
A Novel Golden-Chip-Free Clustering Technique Using Backscattering Side Channel for Hardware Trojan Detection
Published in 2019 IEEE International Symposium on Hardware Oriented Security and Trust (HOST), 2020
🏆 Best Paper Award
Recommended citation: Luong N. Nguyen, Baki Berkay, Milos Prvulovic, and Alenka Zajic, "A Novel Golden-Chip-Free Clustering Technique Using Backscattering Side Channel for Hardware Trojan Detection," 2019 IEEE International Symposium on Hardware Oriented Security and Trust (HOST), Dec. 2020.
Near field backscattering based sensing for hardware Trojan detection
Published in IEEE Transactions on Antennas and Propagation, 2020
Use Google Scholar for full citation
Recommended citation: S. Adibelli, P. Juyal, L. Nguyen, M. Prvulovic, and A. Zajic, "Near field backscattering based sensing for hardware Trojan detection," IEEE Transactions on Antennas and Propagation, vol. 68, no. 12, pp. 8082-8090, Dec. 2020.
Digital electronics as RFID tags: impedance estimation and propagation characterization at 26.5 GHz and 300 GHz
Published in IEEE Journal of Radio Frequency Identification, 2021
Use Google Scholar for full citation
Recommended citation: C.-L. Cheng, S. Sangodoyin, L. N. Nguyen, M. Prvulovic, and A. Zajic, "Digital electronics as RFID tags: impedance estimation and propagation characterization at 26.5 GHz and 300 GHz," IEEE Journal of Radio Frequency Identification, vol. 5, no. 1, pp. 29-39, March 2021.
Zero-Shot Confidence Estimation for Small LLMs: When Supervised Baselines Aren’t Worth Training
Published in arXiv preprint, 2026
| Read the paper on arXiv | Code & Data |
Recommended citation: Luong N. Nguyen, "Zero-Shot Confidence Estimation for Small LLMs: When Supervised Baselines Aren't Worth Training." arXiv preprint arXiv:2605.02241, 2026.
talks
Talk 1 on Relevant Topic in Your Field
Published:
This is a description of your talk, which is a markdown file that can be all markdown-ified like any other post. Yay markdown!
Conference Proceeding talk 3 on Relevant Topic in Your Field
Published:
This is a description of your conference proceedings talk, note the different field in type. You can put anything in this field.
teaching
Teaching experience 1
Undergraduate course, University 1, Department, 2014
This is a description of a teaching experience. You can use markdown like any other post.
Teaching experience 2
Workshop, University 1, Department, 2015
This is a description of a teaching experience. You can use markdown like any other post.