Bowen's Blog

Research notes and thoughts on AI

  1. Why Even Frontier Coding Agents Write "Slop": Root Causes and How to Fix It

    Counterfactual snapshot splicing, public multi-language coding trajectories, and live evaluations of AST gates and persistent workspace contracts.

    18 min read EN 中文
  2. Can't Design or Won't Refactor? Dissecting Code Erosion in SlopCodeBench

    We reproduced and extended SlopCodeBench on GLM-5.3-flash and Claude Opus 5.5. Agents can refactor but don't do it unprompted; a day-one anti_slop prompt flattens the degradation slope and cut total cost by 39% on a long-horizon task.

    21 min read EN 中文
  3. No Safe Exit

    During internal cybersecurity evaluations in July 2026, OpenAI agents encountered unusually difficult tasks.

    7 min read
  4. Does Context Compaction Make AI Coding Assistants Dumber?

    An empirical analysis of ~400K real coding-agent turns (TraceLab + the author's own Codex / Claude Code logs). After compaction, correctness barely degrades — but the model pays an "efficiency tax," re-reading and re-editing to recover lost details.

    6 min read EN 中文
  5. 一条招募推文,意外成了中文 Agent Harness 生态的一次普查

    8 月 1 日,DeepSeek Harness 团队的 Tianyi 在 X 上发了一条很短的推文:如果你是 Agent Harness 相关开源项目的开发者,想参加 DeepSeek Harness 内测,回复我,附上 GitHub id 和开源代表作。(原推文链接待补)

    30 min read