Backtrader-Bench：基于自生成多选题的算法交易 LLM 智能体基准
原标题：Backtrader-Bench: Benchmarking LLM Agents on Algorithmic Trading with Self-Generated MCQs
AI 摘要
Backtrader-Bench 是一个用于评估 LLM 智能体在算法交易中能力的基准框架，包含确定性 MCQ 管道和生成器-求解器过滤管道，可自动生成并验证问题。在 30 道精选题目上，工具增强的智能体（如 GPT-5.5 和 Opus 4.7）单次准确率达 90.0%，比最佳无工具基线（73.0%）高出 17 个百分点。该框架旨在生成训练语料，用于强化学习，最终构建专门的量化交易智能体。
正文节选
Backtrader-Bench: Benchmarking LLM Agents on Algorithmic Trading with Self-Generated MCQs Abstract Evaluating LLM coding agents in algorithmic trading is difficult because static benchmarks risk data contamination and numerical backtest outputs require ground truth from actual code execution. We present Backtrader-Bench, a framework with two complementary pipelines. A deterministic multiple-choice question (MCQ) pipeline generates questions from backtest configurations across five trading strate