Pengfei Gao

Pengfei Gao

Auto-generated profile, based on publicly available information

Pengfei Gao

Bytedance, Beijing, China

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 June 2025
Verified
Conference Paper
Open access
XWXinchen Wang
Xinchen Wang
PGPengfei Gao
Pengfei Gao
XMXiangxin Meng
Xiangxin Meng
CPChao Peng
Chao Peng
RHRuida Hu
Ruida Hu
CGCuiyun Gao
Cuiyun Gao

13 July 2020 in IEEE Transactions on Software Engineering
Verified
Journal Article
Closed access
GPGao Pengfei
Gao Pengfei
XHXie Hongyi
Xie Hongyi
PSPu Sun
Pu Sun
JZJun Zhang
Jun Zhang
FSFu Song
Fu Song
TCTaolue Chen
Taolue Chen
Cryptographic algorithms are widely used to protect data privacy in many aspects of daily lives. Unfortunately, programs implementing cryptographic algorithms may be vulnerable to practical power side-channel attacks, which may infer private data via statistical analysis. To thwart these attacks, several masking schemes have been proposed, giving rise to effective countermeasures for reducing the statistical correlation between private data and power consumptions. However, programs that rely on secure masking schemes are not secure a priori. Indeed, designing effective masking programs is a labor intensive and error-prone task. Although some techniques have been proposed for formally verifying masking countermeasures and for quantifying masking strength, they are currently limited to Boolean programs and suffer from low accuracy. In this work, we propose an approach for formally verifying masking countermeasures of arithmetic programs. Our approach is more accurate for arithmetic programs and more scalable for Boolean programs comparing to the existing approaches. It is essentially a synergistic integration of type inference and model-counting based methods, armed with domain specific heuristics. The type inference system allows a fast deduction of leakage-freeness of most intermediate computations, the model-counting based methods accounts for completeness, namely, to eliminate spurious flaws, and the heuristics facilitate both type inference and model-counting based reasoning, which improve scalability and efficiency in practice. In case that the program does contain leakage, we provide a method to quantify its masking strength. A distuiguished feature of our type sytem lies in its support of compositonal reasoning when verifying programs with procedure calls, so the need of inlining procedures can be significantly reduced. We have implemented our methods in a verification tool QMVERIF which has been extensively evaluated on cryptographic benchmarks including full AES, DES and MAC-Keccak. The experimental results demonstrate the effectiveness and efficiency of our approach, especially for compositional reasoning. In particular, our tool is able to automatically prove leakage-freeness of arithmetic programs for which only manual proofs exist so far; it is also significantly faster than the state-of-the-art tools: EasyCrypt on common arithmetic programs, QMSINFER, SC Sniffer and maskVerif on Boolean programs.

30 June 2026 in Proceedings of the ACM on Software Engineering
Verified
Journal Article
Open access
YXYuan-An Xiao
Yuan-An Xiao
PGPengfei Gao
Pengfei Gao
CPChao Peng
Chao Peng
YXYingfei Xiong
Yingfei Xiong
Multi-turn agent systems based on Large Language Models (LLMs) have become increasingly popular for software engineering tasks. While LLM agents demonstrate promising effectiveness, the high computational cost of input tokens due to ever-growing trajectories remains a significant efficiency concern. Efficiency has been largely overlooked in existing studies and agent products, and this paper addresses this gap by introducing an inference-time trajectory reduction approach that reduces computational costs. By analyzing existing agent trajectories, we demonstrate that useless, redundant, and expired information is widespread across trajectories. Such waste can be identified and reduced without compromising the agent's performance. We propose a simple yet effective trajectory reduction approach, AgentDiet, which automatically removes such waste during agent execution. We implement AgentDiet on a top-performing coding agent, and our evaluation on two LLMs and two benchmarks shows that AgentDiet can reduce input tokens by 39.9%-59.7% and the total computational cost by 21.1%-35.9%, while maintaining the same agent performance. These results indicate that inference-time trajectory reduction is a promising direction for agent systems.

20 September 2020
Verified
Conference Paper
Closed access
YJYuanyuan Ji
Yuanyuan Ji
WZWei Zheng
Wei Zheng
HDHairong Dong
Hairong Dong
PGPengfei Gao
Pengfei Gao
Although trains are more efficient and convenient than other transportation, delays often occur. Accurately predicting the delay time of trains is of great significance to both dispatchers and passengers. The method for predicting the arrival delay time of trains is based on feature selection algorithm and machine learning. First, we collect train delay cases to sort out the delay factors. In addition to internal factors, external factors such as weather and signal failure are also considered. Then, an improved max-relevance and min-redundancy method (mRMR) is used for feature selection. Finally, we apply the method of weighted random forest (wRF) to predict the delay time. The results demonstrate that the feature selection algorithm has a prominent effect on improving the accuracy of the model, and the mean square error based on the weighted random forest has an improvement potential in forecast precision.

9 April 2025 in Proceedings of the ACM on Programming Languages
Verified
Journal Article
Open access
YZYedi Zhang
Yedi Zhang
LHLei Huang
Lei Huang
PGPengfei Gao
Pengfei Gao
FSFu Song
Fu Song
JSJun Sun
Jun Sun
JDJin Song Dong
Jin Song Dong
In the rapidly evolving landscape of neural network security, the resilience of neural networks against bit-flip attacks (i.e., an attacker maliciously flips an extremely small amount of bits within its parameter storage memory system to induce harmful behavior), has emerged as a relevant area of research. Existing studies suggest that quantization may serve as a viable defense against such attacks. Recognizing the documented susceptibility of real-valued neural networks to such attacks and the comparative robustness of quantized neural networks (QNNs), in this work, we introduce BFAVerifier, the first verification framework designed to formally verify the absence of bit-flip attacks against QNNs or to identify all vulnerable parameters in a sound and rigorous manner. BFAVerifier comprises two integral components: an abstraction-based method and an MILP-based method. Specifically, we first conduct a reachability analysis with respect to symbolic parameters that represent the potential bit-flip attacks, based on a novel abstract domain with a sound guarantee. If the reachability analysis fails to prove the resilience of such attacks, then we encode this verification problem into an equivalent MILP problem which can be solved by off-the-shelf solvers. Therefore, BFAVerifier is sound, complete, and reasonably efficient. We conduct extensive experiments, which demonstrate its effectiveness and efficiency across various activation functions, quantization bit-widths, and adversary capabilities.

16 October 2023 in Proceedings of the ACM on Programming Languages
Verified
Journal Article
Open access
PGPengfei Gao
Pengfei Gao
YZYedi Zhang
Yedi Zhang
FSFu Song
Fu Song
TCTaolue Chen
Taolue Chen
FSFrancois-Xavier Standaert
Francois-Xavier Standaert
Masking is one of the most effective countermeasures for securely implementing cryptographic algorithms against power side-channel attacks, the design of which however turns out to be intricate and error-prone. While techniques have been proposed to rigorously verify implementations of cryptographic algorithms, currently they are limited in scalability. To address this issue, compositional approaches have been investigated, but insofar they fail to prove the security of recent efficient implementations. To fill this gap, we propose a novel compositional verification approach. In particular, we introduce two new language-level security notions based on which we propose composition strategies and verification algorithms. Our approach is able to prove efficient implementations, which cannot be done by prior compositional approaches. We implement our approach as a tool CONVINCE and conduct extensive experiments to confirm its efficacy. We also use CONVINCE to further explore the design space of the AES Sbox with least refreshing by replacing its implementation for finite-field multiplication with more efficient counterparts. We automatically prove leakage-freeness of these new versions. As a result, we can effectively reduce 1,600 randomness and 3,200 XOR-operations of the state-of-the-art AES implementation.