Event Signature Transfer: Model-Agnostic Forecast Scenario Construction from Historical Events
研究通过引入无需训练的事件签名转移方法,解决已知事件即将发生但其影响形状、大小或时机未知的问题,将历史事件转换为明确的预测情景。
研究通过引入无需训练的事件签名转移方法,解决已知事件即将发生但其影响形状、大小或时机未知的问题,将历史事件转换为明确的预测情景。
研究发现并诊断了BatchNorm架构中机器遗忘评估的未记录混淆问题,通过权重保留固定点操作方法区分测量失败与编码器失败。
研究通过引入GeoAgent,利用具身导航改进视觉语言模型在地理位置定位中的表现,解决了静态图像分析的局限性。
研究通过文本扰动和归因分析,探讨了多模态时间序列预测中模型是否对文本语义敏感,发现文本内容并非提升预测性能的关键因素。
This work addresses the vulnerability of reward models in language model alignment to “reward hacking,” which can yield superficially high scores without genuine improvements in response quality. The study introduces a novel covariance-geometric perspective to analyze ensemble evaluators, revealing an intrinsic link between common-mode errors and evaluator disagreement, and formally demonstrates that internal scoring alone cannot detect such errors. To mitigate this, the authors propose a dual-anchor Bernstein error bound, integrating sub-Gaussian joint modeling, covariance decomposition, and adaptive search under conditional calibration, yielding a search error bound proportional to √log K. The theoretical claims are validated through 120 Gaussian stress tests and real-model audits, which also expose the limitations of disagreement-based metrics under strong search pressure.
研究通过引入无需训练的事件签名转移方法,解决已知事件即将发生但其影响形状、大小或时机未知的问题,将历史事件转换为明确的预测情景。
研究发现并诊断了BatchNorm架构中机器遗忘评估的未记录混淆问题,通过权重保留固定点操作方法区分测量失败与编码器失败。
研究通过引入GeoAgent,利用具身导航改进视觉语言模型在地理位置定位中的表现,解决了静态图像分析的局限性。
研究通过文本扰动和归因分析,探讨了多模态时间序列预测中模型是否对文本语义敏感,发现文本内容并非提升预测性能的关键因素。
This work addresses the vulnerability of reward models in language model alignment to “reward hacking,” which can yield superficially high scores without genuine improvements in response quality. The study introduces a novel covariance-geometric perspective to analyze ensemble evaluators, revealing an intrinsic link between common-mode errors and evaluator disagreement, and formally demonstrates that internal scoring alone cannot detect such errors. To mitigate this, the authors propose a dual-anchor Bernstein error bound, integrating sub-Gaussian joint modeling, covariance decomposition, and adaptive search under conditional calibration, yielding a search error bound proportional to √log K. The theoretical claims are validated through 120 Gaussian stress tests and real-model audits, which also expose the limitations of disagreement-based metrics under strong search pressure.