How effective are traditional test criteria at detecting bugs in large language models generated code?

• URL: http://arxiv.org/abs/2609.09315v1
• Date: Tue, 08 Sep 2026 18:04:21 GMT
• Title: How effective are traditional test criteria at detecting bugs in large language models generated code?
• Abstract summary: Most faults introduced by Large Language Models are relatively trivial to catch.<n>Fault detection rates remain extremely low, often near zero.<n> prompt-aware oracles can improve fault detection, but their overall effectiveness remains limited.
• Score: 5.2762490553598065
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Test adequacy criteria are widely used to evaluate and guide software testing. Although prior research has extensively examined these criteria using human-written programs, faults, and tests, the increasing adoption of Large Language Models (LLMs) for code generation raises important questions about their effectiveness in detecting LLM-induced faults. To investigate this, we conduct an empirical study involving 5 LLMs and 4 benchmarks, simulating end-to-end workflows in which both code and tests are automatically generated. We collect 6,000+ faulty program instances and evaluate the effectiveness and efficiency of 3 widely used adequacy criteria: statement coverage, branch coverage, and mutation testing. Our findings reveal several key insights. First, most faults introduced by LLMs are relatively trivial to catch. Second, the challenging faults are difficult to trigger using either traditional coverage-based or mutation-based criteria. Third, actual fault detection rates remain extremely low, often near zero, because test oracles fail to capture faulty behavior triggered by the generated test prefixes, exposing a critical limitation of automated test generation. Fourth, prompt-aware oracles can improve fault detection, but their overall effectiveness remains limited, highlighting the need for users to manually reason about test assertions. We further observe that mutation testing only marginally outperforms traditional coverage criteria in both triggering and detecting faults, raising questions about whether its significantly higher application cost is justified in this context.

Related papers

• Do Coverage and Mutation Scores of LLM-Generated Test Suites Correlate with Their Effectiveness? (Replicability Study) [6.228248513679207]
  Studies show that correlations among coverage, mutation, and real-bug detection can largely vanish once test suite size is controlled.<n>We show that the usefulness of coverage and mutation is highly context-dependent.<n>We also find little evidence that test suite size is a dominant confounder for correlations among coverage, mutation, and real-bug detection.
  arXiv  Detail & Related papers  (2026-07-24T19:46:30Z)
• Fail-Aware and Explainable Test Oracle Prediction [4.6927139685668315]
  We present FOCAL, an emerging code LLM-based discriminative oracle predictor.<n>It learns from labeled pairs of test prefixes and methods under test, employs losses that emphasize failing cases during training, and grounds its predictions in statement level behavioral evidence.<n>Compared with the baseline method SEER, we substantially improve performance on failing cases for unseen projects.
  arXiv  Detail & Related papers  (2026-07-13T10:04:36Z)
• LiveFMBench: Unveiling the Power and Limits of Agentic Workflows in Specification Generation [75.05397479715576]
  Large language models (LLMs) and agents have shown promising progress, but their true capabilities and failure modes remain unclear.<n>We present the first systematic and contamination-aware study of LLM- and agent-based formal specification generation for C programs.
  arXiv  Detail & Related papers  (2026-05-02T11:31:33Z)
• VIBEPASS: Can Vibe Coders Really Pass the Vibe Check? [46.85901599242161]
  We evaluate two coupled tasks: emphFault-Triggering Test Generation (FT-Test) and emphFault-targeted Program Repair (FPR)<n>We find that fault-targeted reasoning does not scale with general coding ability.<n>Test-guided repair reveals a complementary insight: when self-generated tests successfully witness a fault, the resulting repair matches or outperforms repair guided by externally provided tests, but tests that fail to witness the fault actively degrade repair below unguided baselines.
  arXiv  Detail & Related papers  (2026-03-16T21:14:28Z)
• Code-A1: Adversarial Evolving of Code LLM and Test LLM via Reinforcement Learning [54.95476453942411]
  Reinforcement learning for code generation relies on verifiable rewards from unit test pass rates.<n>Recent self-play methods unify code and test generation in a single model.<n>Code-A1 achieves code generation performance matching or exceeding models trained on human-annotated tests.
  arXiv  Detail & Related papers  (2026-03-16T17:58:13Z)
• ProbeLLM: Automating Principled Diagnosis of LLM Failures [89.44131968886184]
  We propose ProbeLLM, a benchmark-agnostic automated probing framework that elevates weakness discovery from individual failures to structured failure modes.<n>By restricting probing to verifiable test cases and leveraging tool-augmented generation and verification, ProbeLLM grounds failure discovery in reliable evidence.
  arXiv  Detail & Related papers  (2026-02-13T14:33:13Z)
• Can We Classify Flaky Tests Using Only Test Code? An LLM-Based Empirical Study [40.93176986225226]
  Flaky tests yield inconsistent results when they are repeatedly executed on the same code revision.<n>Previous work evaluated approaches to train machine learning models to classify flaky tests based on identifiers in the test code.
  arXiv  Detail & Related papers  (2026-02-05T09:15:09Z)
• Studying the Impact of Early Test Termination Due to Assertion Failure on Code Coverage and Spectrum-based Fault Localization [48.22524837906857]
  This study is the first empirical study on early test termination due to assertion failure.<n>We investigated 207 versions of 6 open-source projects.<n>Our findings indicate that early test termination harms both code coverage and the effectiveness of spectrum-based fault localization.
  arXiv  Detail & Related papers  (2025-04-06T17:14:09Z)
• Large-scale, Independent and Comprehensive study of the power of LLMs for test case generation [11.517293765116307]
  Unit testing is essential for software reliability, yet manual test creation is time-consuming and often neglected.<n>This study presents the first large-scale empirical evaluation of LLM-generated unit tests at the class level.
  arXiv  Detail & Related papers  (2024-06-28T20:38:41Z)
• Test Oracle Automation in the era of LLMs [52.69509240442899]
  Large Language Models (LLMs) have demonstrated remarkable proficiency in tackling diverse software testing tasks. This paper aims to enable discussions on the potential of using LLMs for test oracle automation, along with the challenges that may emerge during the generation of various types of oracles.
  arXiv  Detail & Related papers  (2024-05-21T13:19:10Z)
• Effective Test Generation Using Pre-trained Large Language Models and Mutation Testing [13.743062498008555]
  We introduce MuTAP for improving the effectiveness of test cases generated by Large Language Models (LLMs) in terms of revealing bugs. MuTAP is capable of generating effective test cases in the absence of natural language descriptions of the Program Under Test (PUTs) Our results show that our proposed method is able to detect up to 28% more faulty human-written code snippets.
  arXiv  Detail & Related papers  (2023-08-31T08:48:31Z)
• Position: AI Evaluation Should Learn from How We Test Humans [65.36614996495983]
  We argue that psychometrics, a theory originating in the 20th century for human assessment, could be a powerful solution to the challenges in today's AI evaluations.
  arXiv  Detail & Related papers  (2023-06-18T09:54:33Z)
• LLMs as Factual Reasoners: Insights from Existing Benchmarks and Beyond [135.8013388183257]
  We propose a new protocol for inconsistency detection benchmark creation and implement it in a 10-domain benchmark called SummEdits. Most LLMs struggle on SummEdits, with performance close to random chance. The best-performing model, GPT-4, is still 8% below estimated human performance.
  arXiv  Detail & Related papers  (2023-05-23T21:50:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.