Pu (Luke) Yi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Pu (Luke) Yi has a total of 16 co-authors. Here are the top co-authors by number of shared publications:
- Sara Achour
Stanford University, Stanford, CA, USA
3 shared publications - Chae Young Lee
Stanford University, Stanford, CA, USA
2 shared publications - Darko Marinov
University of Illinois Urbana-Champaign, Urbana, USA
2 shared publications - Tao Xie
Peking University, Beijing, China
2 shared publications - Anjiang Wei
School of EECS, Peking University, Beijing, China
1 shared publication - Cyrille Artho
KTH Royal Institute of Technology
1 shared publication
Latest Publications
- Open Access
- 3 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 529-542
HyperCam: Low-Power Onboard Computer Vision for IoT Cameras
- 3 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 529-542
Open AccessHyperCam: Low-Power Onboard Computer Vision for IoT Cameras
- Open Access
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 342-360
Early Termination for Hyperdimensional Computing Using Inferential Statistics
Hyperdimensional Computing (HDC) is a brain-inspired, lightweight computing paradigm that has shown great potential for inference on the edge and on emerging hardware technologies, achieving state-of-the-art accuracy on certain classification tasks. HDC classifiers are inherently error resilient and support early termination of inference to approximate classification results. Practitioners have developed heuristic methods to terminate inference early for individual inputs, reducing the computation of inference at the cost of accuracy. These techniques lack statistical guarantees and may unacceptably degrade classification accuracy or terminate inference later than is needed to obtain an accuracy result. We present Omen, the first dynamic HDC optimizer that uses inferential statistics to terminate inference early while providing accuracy guarantees. To realize Omen, we develop a statistical view of HDC that reframes HD computations as statistical sampling and testing tasks, enabling the use of statistical tests. We evaluate Omen on 19 benchmark instantiations of four classification tasks. Omen is computationally efficient, delivering up to 7.21--12.18× inference speed-ups over an unoptimized baseline while only incurring a 0.0--0.7% drop in accuracy. Omen outperforms heuristic methods, achieving an additional 0.04--5.85× inference speed-up over the unoptimized baseline compared to heuristic methods while maintaining higher or comparable accuracy.
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 342-360
Open AccessEarly Termination for Hyperdimensional Computing Using Inferential Statistics
Hyperdimensional Computing (HDC) is a brain-inspired, lightweight computing paradigm that has shown great potential for inference on the edge and on emerging hardware technologies, achieving state-of-the-art accuracy on certain classification tasks. HDC classifiers are inherently error resilient and support early termination of inference to approximate classification results. Practitioners have developed heuristic methods to terminate inference early for individual inputs, reducing the computation of inference at the cost of accuracy. These techniques lack statistical guarantees and may unacceptably degrade classification accuracy or terminate inference later than is needed to obtain an accuracy result. We present Omen, the first dynamic HDC optimizer that uses inferential statistics to terminate inference early while providing accuracy guarantees. To realize Omen, we develop a statistical view of HDC that reframes HD computations as statistical sampling and testing tasks, enabling the use of statistical tests. We evaluate Omen on 19 benchmark instantiations of four classification tasks. Omen is computationally efficient, delivering up to 7.21--12.18× inference speed-ups over an unoptimized baseline while only incurring a 0.0--0.7% drop in accuracy. Omen outperforms heuristic methods, achieving an additional 0.04--5.85× inference speed-up over the unoptimized baseline compared to heuristic methods while maintaining higher or comparable accuracy.
- Open Access
- 5 April 2024
- Published by Springer Nature
- p. 3-22
JPF: From 2003 to 2023
We give an account of JPF’s current architecture as it has evolved over the last 20 years. Key changes include a modular, extensible design, and Java 11 support.Java 11 brought with it fundamental changes in the language and its runtime, in particular, a new modular library system, different compilation of string expressions to bootstrap methods, and changes in many internal interfaces that allow access to the loaded code and the virtual machine state. These changes required numerous adaptations in JPF to ensure a successful compilation and correct behavior under Java 11.
- 5 April 2024
- Published by Springer Nature
- p. 3-22
Open AccessJPF: From 2003 to 2023
We give an account of JPF’s current architecture as it has evolved over the last 20 years. Key changes include a modular, extensible design, and Java 11 support.Java 11 brought with it fundamental changes in the language and its runtime, in particular, a new modular library system, different compilation of string expressions to bootstrap methods, and changes in many internal interfaces that allow access to the loaded code and the virtual machine state. These changes required numerous adaptations in JPF to ensure a successful compilation and correct behavior under Java 11.
- Open Access
- 16 October 2023
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Programming Languages
- Vol. 7 (OOPSLA2) , 1-30
Hardware-Aware Static Optimization of Hyperdimensional Computations
Binary spatter code (BSC)-based hyperdimensional computing (HDC) is a highly error-resilient approximate computational paradigm suited for error-prone, emerging hardware platforms. In BSC HDC, the basic datatype is a hypervector , a typically large binary vector, where the size of the hypervector has a significant impact on the fidelity and resource usage of the computation. Typically, the hypervector size is dynamically tuned to deliver the desired accuracy; this process is time-consuming and often produces hypervector sizes that lack accuracy guarantees and produce poor results when reused for very similar workloads. We present Heim, a hardware-aware static analysis and optimization framework for BSC HD computations. Heim analytically derives the minimum hypervector size that minimizes resource usage and meets the target accuracy requirement. Heim guarantees the optimized computation converges to the user-provided accuracy target on expectation, even in the presence of hardware error. Heim deploys a novel static analysis procedure that unifies theoretical results from the neuroscience community to systematically optimize HD computations. We evaluate Heim against dynamic tuning-based optimization on 25 benchmark data structures. Given a 99% accuracy requirement, Heim-optimized computations achieve a 99.2%-100.0% median accuracy, up to 49.5% higher than dynamic tuning-based optimization, while achieving 1.15x-7.14x reductions in hypervector size compared to HD computations that achieve comparable query accuracy and finding parametrizations 30.0x-100167.4x faster than dynamic tuning-based approaches. We also use Heim to systematically evaluate the performance benefits of using analog CAMs and multiple-bit-per-cell ReRAM over conventional hardware, while maintaining iso-accuracy – for both emerging technologies, we find usages where the emerging hardware imparts significant benefits.
- 16 October 2023
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Programming Languages
- Vol. 7 (OOPSLA2) , 1-30
Open AccessHardware-Aware Static Optimization of Hyperdimensional Computations
Binary spatter code (BSC)-based hyperdimensional computing (HDC) is a highly error-resilient approximate computational paradigm suited for error-prone, emerging hardware platforms. In BSC HDC, the basic datatype is a hypervector , a typically large binary vector, where the size of the hypervector has a significant impact on the fidelity and resource usage of the computation. Typically, the hypervector size is dynamically tuned to deliver the desired accuracy; this process is time-consuming and often produces hypervector sizes that lack accuracy guarantees and produce poor results when reused for very similar workloads. We present Heim, a hardware-aware static analysis and optimization framework for BSC HD computations. Heim analytically derives the minimum hypervector size that minimizes resource usage and meets the target accuracy requirement. Heim guarantees the optimized computation converges to the user-provided accuracy target on expectation, even in the presence of hardware error. Heim deploys a novel static analysis procedure that unifies theoretical results from the neuroscience community to systematically optimize HD computations. We evaluate Heim against dynamic tuning-based optimization on 25 benchmark data structures. Given a 99% accuracy requirement, Heim-optimized computations achieve a 99.2%-100.0% median accuracy, up to 49.5% higher than dynamic tuning-based optimization, while achieving 1.15x-7.14x reductions in hypervector size compared to HD computations that achieve comparable query accuracy and finding parametrizations 30.0x-100167.4x faster than dynamic tuning-based approaches. We also use Heim to systematically evaluate the performance benefits of using analog CAMs and multiple-bit-per-cell ReRAM over conventional hardware, while maintaining iso-accuracy – for both emerging technologies, we find usages where the emerging hardware imparts significant benefits.
- Open Access
- 30 March 2022
- Published by Springer Nature
- p. 217-235
A Theoretical Analysis of Random Regression Test Prioritization
Regression testing is an important activity to check software changes by running the tests in a test suite to inform the developers whether the changes lead to test failures. Regression test prioritization (RTP) aims to inform the developers faster by ordering the test suite so that tests likely to fail are run earlier. Many RTP techniques have been proposed and are often compared with the random RTP baseline by sampling some of the n! different test-suite orders for a test suite with n tests. However, there is no theoretical analysis of random RTP. We present such an analysis, deriving probability mass functions and expected values for metrics and scenarios commonly used in RTP research. Using our analysis, we revisit some of the most highly cited RTP papers and find that some presented results may be due to insufficient sampling. Future RTP research can leverage our analysis and need not use random sampling but can use our simple formulas or algorithms to more precisely compare with random RTP.
- 30 March 2022
- Published by Springer Nature
- p. 217-235
Open AccessA Theoretical Analysis of Random Regression Test Prioritization
Regression testing is an important activity to check software changes by running the tests in a test suite to inform the developers whether the changes lead to test failures. Regression test prioritization (RTP) aims to inform the developers faster by ordering the test suite so that tests likely to fail are run earlier. Many RTP techniques have been proposed and are often compared with the random RTP baseline by sampling some of the n! different test-suite orders for a test suite with n tests. However, there is no theoretical analysis of random RTP. We present such an analysis, deriving probability mass functions and expected values for metrics and scenarios commonly used in RTP research. Using our analysis, we revisit some of the most highly cited RTP papers and find that some presented results may be due to insufficient sampling. Future RTP research can leverage our analysis and need not use random sampling but can use our simple formulas or algorithms to more precisely compare with random RTP.
- Open Access
- 20 March 2021
- Published by Springer Nature
- Vol. 12651, 270-287
Probabilistic and Systematic Coverage of Consecutive Test-Method Pairs for Detecting Order-Dependent Flaky Tests
Software developers frequently check their code changes by running a set of tests against their code. Tests that can nondeterministically pass or fail when run on the same code version are called flaky tests. These tests are a major problem because they can mislead developers to debug their recent code changes when the failures are unrelated to these changes. One prominent category of flaky tests is order-dependent (OD) tests, which can deterministically pass or fail depending on the order in which the set of tests are run. By detecting OD tests in advance, developers can fix these tests before they change their code. Due to the high cost required to explore all possible orders (n! permutations for n tests), prior work has developed tools that randomize orders to detect OD tests. Experiments have shown that randomization can detect many OD tests, and that most OD tests depend on just one other test to fail. However, there was no analysis of the probability that randomized orders detect OD tests. In this paper, we present the first such analysis and also present a simple change for sampling random test orders to increase the probability. We finally present a novel algorithm to systematically explore all consecutive pairs of tests, guaranteeing to detect all OD tests that depend on one other test, while running substantially fewer orders and tests than simply running all test pairs.
- 20 March 2021
- Published by Springer Nature
- Vol. 12651, 270-287
Open AccessProbabilistic and Systematic Coverage of Consecutive Test-Method Pairs for Detecting Order-Dependent Flaky Tests
Software developers frequently check their code changes by running a set of tests against their code. Tests that can nondeterministically pass or fail when run on the same code version are called flaky tests. These tests are a major problem because they can mislead developers to debug their recent code changes when the failures are unrelated to these changes. One prominent category of flaky tests is order-dependent (OD) tests, which can deterministically pass or fail depending on the order in which the set of tests are run. By detecting OD tests in advance, developers can fix these tests before they change their code. Due to the high cost required to explore all possible orders (n! permutations for n tests), prior work has developed tools that randomize orders to detect OD tests. Experiments have shown that randomization can detect many OD tests, and that most OD tests depend on just one other test to fail. However, there was no analysis of the probability that randomized orders detect OD tests. In this paper, we present the first such analysis and also present a simple change for sampling random test orders to increase the probability. We finally present a novel algorithm to systematically explore all consecutive pairs of tests, guaranteeing to detect all OD tests that depend on one other test, while running substantially fewer orders and tests than simply running all test pairs.