In recent years, the practice of fuzzing Deep Learning (DL) APIs has received significant attention in the software engineering community. Many API-level DL fuzzers have been proposed to test individual DL APIs by generating malformed input. Although these fuzzers have been effective in detecting bugs and outperforming prior work, there remains a gap in benchmarking them against ground-truth, real-world bugs in DL libraries. Existing comparisons among these API-level DL fuzzers primarily focus on the bugs detected but do not offer a comprehensive, in-depth evaluation of the fuzzers’ effectiveness. In this work, we perform the first in-depth evaluation of state-of-the-art API-level DL fuzzers that generate tests for single DL APIs, focusing on their effectiveness against real-world bugs. We manually created an extensive benchmark dataset, including 517 real-world DL bugs collected from PyTorch and TensorFlow libraries that can be triggered by malformed inputs. We then apply seven state-of-the-art DL fuzzers— FreeFuzz , DeepRel , NablaFuzz , DocTer , ACETest , TitanFuzz , and FuzzGPT —to our benchmark dataset, following their respective instructions. Our results show that these fuzzers detect only 6.5% (34 out of 517) of the unique real-world bugs in the dataset. Our analysis identifies two dominant factors that impact the effectiveness of these fuzzers in detecting real-world bugs. These findings suggest opportunities for improving the performance of fuzzers in future work. Overall, this study extends previous work on DL fuzzers by providing an extensive evaluation and benchmarking platform for fuzzing DL libraries.

Program slicing is a critical technique in software engineering, enabling developers to isolate relevant portions of code for tasks such as bug detection, code comprehension, and debugging. In this study, we investigate the application of large language models (LLMs) to both static and dynamic program slicing, with a focus on Java programs. We evaluate the performance of four state-of-the-art LLMs, i.e., GPT-4o, GPT-3.5 Turbo, Llama-2, and Gemma-7B, by leveraging advanced prompting techniques, including few-shot learning and chain-of-thought reasoning. Using a dataset of 100 Java programs derived from LeetCode problems, our experiments reveal that GPT-4o performs the best in both static and dynamic slicing across other LLMs, achieving an accuracy of 60.84% and 59.69%, respectively. Our results also show that the LLMs we experimented with are yet to achieve reasonable performance for either static slicing or dynamic slicing. Through a rigorous manual analysis, we developed a taxonomy of root causes and failure locations to explore the unsuccessful cases in more depth. We identified Complex Control Flow as the most frequent root cause of failures, with the majority of issues occurring in Variable Declarations and Assignments locations. To improve the performance of LLMs, we further examined a widely-used strategy for prompting guided by our taxonomy, i.e., prompt crafting, which involved refining the prompts to better guide the LLM through the slicing process. Our evaluation shows that prompt crafting can improve accuracy by 4%.

Recently, many Deep Learning (DL) fuzzers have been proposed for API-level testing of DL libraries. However, they either perform unguided input generation (e.g., not considering the relationship between API arguments when generating inputs) or only support a limited set of corner-case test inputs. Furthermore, many developer APIs crucial for library development remain untested, as they are typically not well documented and lack clear usage guidelines, unlike end-user APIs. This makes them a more challenging target for automated testing. To fill this gap, we propose a novel fuzzer named Orion, which combines guided test input generation and corner-case test input generation based on a set of fuzzing heuristic rules constructed from historical data known to trigger critical issues in the underlying implementation of DL APIs. To extract the fuzzing heuristic rules, we first conduct an empirical study on the root cause analysis of 376 vulnerabilities in two of the most popular DL libraries, PyTorch and TensorFlow. We then construct the fuzzing heuristic rules based on the root causes of the extracted historical vulnerabilities. Using these fuzzing heuristic rules, Orion generates corner-case test inputs for API-level fuzzing. In addition, we extend the seed collection of existing studies to include test inputs for developer APIs. Our evaluation shows that Orion reports 135 vulnerabilities in the latest releases of TensorFlow and PyTorch, 76 of which were confirmed by the library developers. Among the 76 confirmed vulnerabilities, 69 were previously unknown, and 7 have already been fixed. The rest are awaiting further confirmation. For end-user APIs, Orion detected 45.58% and 90% more vulnerabilities in TensorFlow and PyTorch, respectively, compared to the state-of-the-art conventional fuzzer, DeepRel. When compared to the state-of-the-art LLM-based DL fuzzer, AtlasFuz, and Orion detected 13.63% more vulnerabilities in TensorFlow and 18.42% more vulnerabilities in PyTorch. Regarding developer APIs, Orion stands out by detecting 117% more vulnerabilities in TensorFlow and 100% more vulnerabilities in PyTorch compared to the most relevant fuzzer designed for developer APIs, such as FreeFuzz.