NeurIPS

FEEL: Quantifying Heterogeneity in Physiological Signals for Generalizable Emotion Recognition

FEEL：量化生理信号异质性以实现通用情感识别

Pragya Singh, Ankush Gupta, Somay Jalan, Mohan Kumar, Pushpendra Singh

AI总结 FEEL通过评估19个数据集上的EDA和PPG信号，探讨了情感识别中异质性的影响，发现CLSP模型在F1分数上表现最佳，而传统模型和手工特征方法也具有竞争力。

Comments Published at Conference on Neural Information Processing Systems (NeurIPS 2025) Track on Datasets and Benchmarks

Modelling Cascading Physical Climate Risk in Supply Chains with Adaptive Firms: A Spatial Agent-Based Framework

基于适应性企业的供应链中物理气候风险 cascading 的建模：一种空间 agent-based 框架

Yara Mohajerani

机构 * Quantile Labs 纠错

AI总结 本文提出一个开源 Python 框架，用于建模供应链经济中的 cascading 物理气候风险，通过整合地理洪水风险与企业家庭 agent-based 模型，模拟直接资产损失和间接中断。企业通过资本硬化和备用供应商搜索适应，减少损失和中断，强调间接级联效应的重要性。

Comments V1 presented at NeurIPS 2025 Tackling Climate Change with Machine Learning workshop. V4 replaces evolutionary learning with explicit firm continuity adaptation, adds stock-flow consistency, matched-seed ensembles, cascade diagnostics, and internal validations. Code: https://github.com/yaramohajerani/spatial-climate-ABM

TS-Agent: Understanding and Reasoning Over Raw Time Series via Iterative Insight Gathering

TS-Agent：通过迭代洞察获取理解并推理原始时间序列

Penghang Liu, Elizabeth Fons, Annita Vapsi, Mohsen Ghassemi, Svitlana Vyetrenko, Daniel Borrajo, Vamsi K. Potluru, Manuela Veloso

机构 * JPMorgan AI Research, NY, USA（摩根大通人工智能研究院，纽约，美国） 纠错

AI总结 TS-Agent通过迭代洞察获取机制，在时间序列理解和推理任务中匹配或超越文本、视觉及时间序列语言模型基线，尤其在零样本设置下表现更优。

Comments NeurIPS 2025 Workshop on Foundations of Reasoning in Language Models