| 年份 | 论文 | 主题 | 错误类型 | 作者承认(节录) |
|---|---|---|---|---|
| 2022 | A Novel Convergence Analysis for the Stochastic Proximal Point Alg… | 优化理论 | 循环论证/未证断言 | No, Lemma 1 can't be fixed. The result simply does not hold… |
| 2024 | Reward Adaptation Via Q-Manipulation | 强化学习理论 | 错误常数/系数/速率 | “we thank the reviewers for pointing out the error in Lemma 5.”… |
| 2024 | Farzi Data: Autoregressive Data Distillation | 生成模型/采样 | 缺失/过弱假设 | Thank you for catching our honest mistake at the end of Theorem 3.1… |
| 2018 | From Information Bottleneck To Activation Norm Penalty | 优化理论 | 循环论证/未证断言 | As you rightly pointed out, the lack of treatment for the second term (log-determinant) is unsatisfying… |
| 2020 | Policy Optimization with Stochastic Mirror Descent | 强化学习理论 | 循环论证/未证断言 | The reviewer's comment of Eq.(40) is correct… |
| 2020 | An Information Theoretic Perspective on Disentangled Representatio… | 概率/集中不等式 | 缺失/过弱假设 | As you point out, now we realize that the factorized noise is sufficient for conditional independence… |
| 2018 | Residual Loss Prediction: Reinforcement Learning With No Incremen… | 强化学习理论 | 缺失/过弱假设 | indeed, you're right, there's a missing term. In going from Eq 7 to Eq 8… |
| 2018 | Three factors influencing minima in SGD | 优化理论 | 缺失/过弱假设 | We agree there is a mathematical mistake in allowing sigma to vary with theta… |
| 2023 | Shuffle Gaussian Mechanism for Differential Privacy | 概率/集中不等式 | 量词/作用域错误 | Thanks for the insightful comment! We have indeed overlooked the fact that normalization and shuffling… |
| 2023 | Improving Out-of-distribution Generalization with Indirection Repr… | 线性代数/谱 | 无效放缩/不等式步骤 | In the revision, we provided the definition for the ‖·‖∞ norm in Definition A.3… |
| 2024 | Learning multi-modal generative models with permutation-invariant… | 生成模型/采样 | 缺失/过弱假设 | We agree that Proposition 1 relies on a strong assumption about the encoding distribution… |
| 2024 | Are Transformers with One Layer Self-Attention Using Low-Rank Weig… | 学习理论/泛化 | 无效放缩/不等式步骤 | As for the bugs in equation (10), you are absolutely right… |