EvoProbe: Evaluation-Driven Benchmark Evolution for LLM Agents

arXiv:2610.19433 · cs.AI · Submitted 2026-10-14 · Read on arXiv

cs.AI

Submitted: 2026-10-14

Updated: 2026-10-14

Code: https://github.com/Qhovo1/EvoProbe

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

Terminology

• Benchmark — 1× in this paper · explained in 2 episodes

Sources

• Where Did It Go Wrong? Process-Level Evaluation of Web Agents with Semantic State Tracking
• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
• ExploitBench: A Capability Ladder Benchmark for LLM Cybersecurity Agents
• ExploitGym: Can AI Agents Turn Security Vulnerabilities into Real Attacks?

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems