Sunny ChokshiAI & Security Researcher

SITE MAP

Every page on this site

A plain list of all pages, posts and papers. Search engines read the XML versions linked at the bottom.

Main pages

  • HomeSunny Chokshi studies how a single comment in source code can fool LLM-based code review, and how to stop it. PhD student, University of the Cumberlands.
  • Research on LLM Security and Code ReviewSunny Chokshi studies adversarial comments, LLM code vulnerability detection, and practical defenses for AI-assisted security review.
  • Publications on LLM SecurityPeer-reviewed papers and preprints by Sunny Chokshi on adversarial comment injection, LLM code security, and cryptographic API misuse.
  • Peer Review and Academic ServicePeer review and conference service by Sunny Chokshi, including IEEE Transactions on Dependable and Secure Computing, IEEE Access, IEEE ICon INDIA, and ICAISG.
  • LLM Security Datasets and CodeOpen datasets, replication packages, and code by Sunny Chokshi for testing LLM code review on adversarial comments and cryptographic API misuse.
  • LLM Security Research NotesSunny Chokshi's research notes on LLM code security, adversarial comments, cryptographic API misuse, and false alarms, with links to papers and datasets.
  • Research ProjectsOpen research projects and tools by Sunny Chokshi: CommentInject, CryptoBench, the Detection Rate replication package, and Does Indirection Hurt?
  • Conference PresentationsSunny Chokshi's IEEE Cyber-AI 2026 presentations on adversarial comment injection and defenses for LLM-based code vulnerability detection.
  • Contact Sunny ChokshiContact Sunny Chokshi about research in LLM security, AI code review, adversarial robustness, and cryptographic API misuse.

Research notes

Papers

For crawlers and readers