EvoVuln evolves executable detection policies for five smart-contract vulnerability types using cold-start synthetic testing followed by few-shot refinement on five vulnerable and five safe contracts, reaching 71% macro F1 and enabling a small model to beat a large zero-shot model by 19 points at un
Chain-of-thought prompting of large language models for discovering and fixing software vulnerabilities
10 Pith papers cite this work. Polarity classification is still indexing.
10
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 10roles
background 1polarities
background 1representative citing papers
LLM-based security code review is vulnerable to framing bias, with a novel iterative refinement attack achieving 100% success in reintroducing vulnerabilities across real projects.
Empirical benchmark RepBench shows AST+PDG yields 83.2% accuracy for LLM vulnerability reasoning vs 53.5% for raw source, with graph-only prompts superior due to reduced context dilution.
MA-CoT prompting reduces security findings in LLM-generated code by 57.6% on a 200-task dataset and 94.5% on LLMSecEval across C, Java, and Python, outperforming vanilla, zero-shot, and standard CoT strategies.
PromptAudit evaluates five prompting strategies across five LLMs on 1000 CVEs and finds chain-of-thought prompting yields the strongest overall performance while adaptive chain-of-thought and self-consistency reduce effective results.
ReasonVul deploys three LLM agents with independent analysis and structured debate to achieve 40% PairAcc and 72.52% F1 on PrimeVul, outperforming baselines by 81% in PairAcc.
A within-subject study of 12 developers found that security training reduced validated weaknesses by 31.5% and critical issues by 79.2% in LLM-assisted backend coding.
MultiVul uses multimodal contrastive learning to align code and comment representations, yielding up to 27% F1 gains on vulnerability detection benchmarks over prompting and code-only baselines.
Empathic similarity feedback in prompts generates more acceptable compromises than chain-of-thought, and margin-based training on the resulting data lets smaller models produce them without ongoing empathy estimation.
VulWeaver improves Java vulnerability detection to 0.75 F1 by enhancing dependency graphs with LLM semantic fixes, extracting full context from slices plus implicit usage info, and applying type-specific meta-prompting with majority voting.
citing papers explorer
-
Knowledge Over Parameters: Evolving Smart Contract Vulnerability Detection
EvoVuln evolves executable detection policies for five smart-contract vulnerability types using cold-start synthetic testing followed by few-shot refinement on five vulnerable and five safe contracts, reaching 71% macro F1 and enabling a small model to beat a large zero-shot model by 19 points at un
-
Measuring and Exploiting Contextual Bias in LLM-Assisted Security Code Review
LLM-based security code review is vulnerable to framing bias, with a novel iterative refinement attack achieving 100% success in reintroducing vulnerabilities across real projects.
-
Representation Matters: An Empirical Study of Program Representations for LLM Vulnerability Reasoning
Empirical benchmark RepBench shows AST+PDG yields 83.2% accuracy for LLM vulnerability reasoning vs 53.5% for raw source, with graph-only prompts superior due to reduced context dilution.
-
Enhancing Reliability in LLM-Based Secure Code Generation
MA-CoT prompting reduces security findings in LLM-generated code by 57.6% on a 200-task dataset and 94.5% on LLMSecEval across C, Java, and Python, outperforming vanilla, zero-shot, and standard CoT strategies.
-
PromptAudit: Auditing Prompt Sensitivity in LLM-Based Vulnerability Detection
PromptAudit evaluates five prompting strategies across five LLMs on 1000 CVEs and finds chain-of-thought prompting yields the strongest overall performance while adaptive chain-of-thought and self-consistency reduce effective results.
-
Three Heads Are Better Than One: A Multi-perspective Reasoning Framework for Enhanced Vulnerability Detection
ReasonVul deploys three LLM agents with independent analysis and structured debate to achieve 40% PairAcc and 72.52% F1 on PrimeVul, outperforming baselines by 81% in PairAcc.
-
A Quasi-Experimental Developer Study of Security Training in LLM-Assisted Web Application Development
A within-subject study of 12 developers found that security training reduced validated weaknesses by 31.5% and critical issues by 79.2% in LLM-assisted backend coding.
-
Learning Generalizable Multimodal Representations for Software Vulnerability Detection
MultiVul uses multimodal contrastive learning to align code and comment representations, yielding up to 27% F1 gains on vulnerability detection benchmarks over prompting and code-only baselines.
-
Generating Place-Based Compromises Between Two Points of View
Empathic similarity feedback in prompts generates more acceptable compromises than chain-of-thought, and margin-based training on the resulting data lets smaller models produce them without ongoing empathy estimation.
-
VulWeaver: Weaving Broken Semantics for Grounded Vulnerability Detection
VulWeaver improves Java vulnerability detection to 0.75 F1 by enhancing dependency graphs with LLM semantic fixes, extracting full context from slices plus implicit usage info, and applying type-specific meta-prompting with majority voting.