这篇论文通过精心设计的安慰剂对照实验，拆解了自我修复反馈中真正起作用的成分——不是重新看到失败代码，而是比对可执行反例。方法严谨，结果清晰。
研究使用安慰剂对照方法分解反馈包，针对0.5B-1.5B的冻结代码模型（共3个）在HumanEval+/MBPP+基准上的290个死任务单元进行测试。主实验生成7000个新样本，盲重采样比纯代码重试多解锁18个（25/7，p=0.0021），代码加事实比纯代码多解锁18个（21/3，p=0.00042），与通用子弹安慰剂相比多15个（p=0.0041）。指令单独效果不明显（+3，p=0.36）。研究结论是反馈价值来自与外部可执行反例的比较，而非重新暴露于失败代码。
Falsification, Not Exposure: An Internally Preregistered Placebo-Controlled Decomposition of Self-Repair Feedback in Frozen Small Code Models
In deployment settings where retraining is infeasible, small frozen code models are routinely asked to repair a failed program after seeing their own failing output, usually treated as a retry mechanism. From a Popperian view, a generated program is a conjecture and a test-execution violation is an oracle-relative, executable counterexample, so feedback's value should be attributed not to re-exposure to failing code but to whether the conjecture is opened to external, executable criticism. As the third stage of a falsification-centered measurement program, this study builds a placebo-controlled instrument that decomposes the feedback packet against a blind-resampling baseline at matched output-generation budget and against content-free, shape-matched placebos. The contribution is not a new repair algorithm but a reflexive methodology (packet decomposition, placebo mirroring, matched-budget discordant-pair tests, fresh-generation confirmation, executable audits) that makes both the model's program conjecture and the researcher's "feedback content works" claim falsifiable. Across six HumanEval+/MBPP+ cells with three 0.5B-1.5B frozen models, 290 dead task-cell units (no best-of-8 candidate passing the public tier) were evaluated; the main run produced 7,000 fresh generations and a preregistered follow-up 1,400 more. Blind resampling exceeded bare-code retry by +18 net unlocks (25/7, Holm p=0.0021). Code-plus-facts recovered +18 over bare code (21/3, p=0.00042) and +15 over a generic-bullet placebo (p=0.0041). An instruction-only effect was not distinguishable (+3, p=0.36). Code-plus-facts and blind resampling tied at 26 unlocks each (not equivalence). Six external-controller follow-ups tied a content-free shape placebo. In this regime, falsification helped not as vocabulary or self-critique, but as comparison with external, executable counterexamples.