让分子采样更快更准，只需几步就能算出可靠概率。
FALCON: Few-step Accurate Likelihoods for Continuous Flows
- 用混合训练目标提升连续流的可逆性，实现少步采样。
- 在分子玻尔兹曼采样上超越现有模型，速度提升100倍。
- 适合需要高效精准采样的分子模拟研究者使用。
在统计物理中，大规模分子状态的热平衡采样是一项长期挑战。玻尔兹曼生成器通过将可精确计算似然的概率生成模型与重要性采样结合，获得目标分布的一致样本。当前主流方法使用连续归一化流（CNF）并采用流匹配进行高效训练，但其似然计算代价极高，每样本需数千次函数求值，严重限制了应用。本文提出少步高精度似然方法（FALCON），通过引入一种混合训练目标，增强连续流的可逆性，实现仅需少量步骤即可获得足够准确的似然，适用于重要性采样。实验表明，FALCON在分子玻尔兹曼采样任务中优于现有最优归一化流模型，且比性能相当的CNF模型快两个数量级。
原文摘要 · Abstract (English)
Scalable sampling of molecular states in thermodynamic equilibrium is a long-standing challenge in statistical physics. Boltzmann Generators tackle this problem by pairing a generative model, capable of exact likelihood computation, with importance sampling to obtain consistent samples under the target distribution. Current Boltzmann Generators primarily use continuous normalizing flows (CNFs) trained with flow matching for efficient training of powerful models. However, likelihood calculation for these models is extremely costly, requiring thousands of function evaluations per sample, severely limiting their adoption. In this work, we propose Few-step Accurate Likelihoods for Continuous Flows (FALCON), a method which allows for few-step sampling with a likelihood accurate enough for importance sampling applications by introducing a hybrid training objective that encourages invertibility. We show FALCON outperforms state-of-the-art normalizing flow models for molecular Boltzmann sampling and is two orders of magnitude faster than the equivalently performing CNF model.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。