Abstract
We leverage recent breakthroughs in neural density estimation to propose anew unsupervised anomaly detection technique (ANODE). By estimating theprobability density of the data in a signal region and in sidebands, andinterpolating the latter into the signal region, a likelihood ratio of data vs.background can be constructed. This likelihood ratio is broadly sensitive tooverdensities in the data that could be due to localized anomalies. Inaddition, a unique potential benefit of the ANODE method is that the backgroundcan be directly estimated using the learned densities. Finally, ANODE is robustagainst systematic differences between signal region and sidebands, giving itbroader applicability than other methods. We demonstrate the power of this newapproach using the LHC Olympics 2020 R\&D Dataset. We show how ANODE canenhance the significance of a dijet bump hunt by up to a factor of 7 with a10\% accuracy on the background prediction. While the LHC is used as therecurring example, the methods developed here have a much broader applicabilityto anomaly detection in physics and beyond.