RECEIPT: Deterministic, Reward-Hacking-Resistant Verification for White-Box Agentic XSS Discovery

arXiv:2607.18575 · cs.CR · Submitted 2026-07-20 · Read on arXiv

Muxi Lyu, Karen Shieh, Yiwei Hou, Hao Wang, Koushik Sen, David Wagner

cs.CR

Submitted: 2026-07-20

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Verifier — 57× in this paper · explained in 2 episodes
• Verification — 39× in this paper · explained in 2 episodes
• Reward Hacking — 18× in this paper · explained in 8 episodes
• Benchmark — 14× in this paper · explained in 2 episodes
• AI Agents — 5× in this paper · explained in 2 episodes
• Ablation — 4× in this paper · explained in 1 episode
• Network and Distributed System Security Symposium — 4× in this paper
• International Conference on Software Engineering — 3× in this paper
• Analyzer — 2× in this paper · explained in 1 episode
• Chrome DevTools Protocol — 2× in this paper
• Grounding — 2× in this paper · explained in 3 episodes
• International Conference on Learning Representations — 2× in this paper
• Actor — 1× in this paper · explained in 3 episodes
• Agentic Discovery — 1× in this paper · explained in 1 episode
• Agentic System — 1× in this paper · explained in 2 episodes
• Core — 1× in this paper · explained in 2 episodes
• Hallucination — 1× in this paper · explained in 3 episodes
• IEEE Symposium on Security and Privacy — 1× in this paper

Sources

• LLM Agents can Autonomously Exploit One-day Vulnerabilities
• EnIGMA: Interactive Tools Substantially Assist LM Agents in Finding Security Vulnerabilities
• LLM Agents can Autonomously Hack Websites
• Revelio: Cost-Efficient Agentic Memory Safety Vulnerability Detection For Repository-Scale Codebases
• Defining and Characterizing Reward Hacking
• Natural Emergent Misalignment from Reward Hacking in Production RL
• Cybench: A Framework for Evaluating Cybersecurity Capabilities and Risks of Language Models
• CyberGym: Evaluating AI Agents' Real-World Cybersecurity Capabilities at Scale
• Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation
• Establishing Best Practices for Building Rigorous Agentic Benchmarks

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs