ProcCtrlBench: Evaluating Process-Level Defects and Control Preservation in LLM Coding Agents 事件

ProcCtrlBench: Evaluating Process-Level Defects and Control Preservation in LLM Coding Agents arXiv:2605.20251v3 Announce Type: replace-cross Abstract: Existing benchmarks for LLM coding agents primarily evaluate final outcomes. While useful for measuring overall capability, these metrics provide limited visibility and often miss defects that arise during execution. We present ProcCtrlBench, a benchmark for execution-process evaluation in LLM coding agents. ProcCtrlBench organizes recurrent exe