Explore the latest research from Facebook

In this work, we highlight an intrinsic correspondence between a simple family of autoregressive normalizing flows and identifiable causal models. We exploit the fact that autoregressive flow architectures define an ordering over variables, analogous to a causal ordering, to show that they are well-suited to performing a range of causal inference tasks, ranging from causal discovery to making interventional and counterfactual predictions.

In our estimation procedure, we introduce a cluster-based regression adjustment that substantially improves precision for estimating global treatment effects, as well as a procedure to test for interference. With our regression adjustment, we find that imbalanced clusters can better account for interference than balanced clusters without sacrificing accuracy.

We present Clockwork, a delay-based global scheduling framework for data pipelines which improves landing time stability by spreading out tasks throughout the day.

We propose a theoretical framework for studying such amplification in a matrix factorization based recommender system. We model the dynamics of the system, where users interact with the recommender systems and gradually “drift” toward the recommended content, with the recommender system adapting, based on user feedback, to the updated preferences.

This paper describes a scalable solution to deploy JupyterHub for computational social science research on the cloud. We use a reference architecture on AWS to walk through the design principles and details.

The cumulative approach avoids binning that smooths over deviations of the subpopulation from the full population. Such cumulative aggregation furnishes both high-resolution graphical methods and simple scalar summary statistics (analogous to those of Kuiper and of Kolmogorov and Smirnov used in statistical significance testing for comparing probability distributions).

We show that it improves on the existing approaches that use the Cramer-Chernoff technique such as the Hoeffding, Bernstein, and Bennett inequalities. The resulting confidence sequence is confirmed to be favorable in synthetic coverage problems, adaptive stopping algorithms, and multi-armed bandit problems.

In this talk, we discuss how these systems will need to evolve from the traditional formulations by incorporating the producer value into their objectives. Jointly optimizing the ranking functions behind these systems on both consumer and producer values is a new direction and raises many technical challenges.

Motivated by a mechanism used in practice by several companies, this paper considers a smoothing procedure that relies on pacing multipliers: on behalf of each buyer, the auction market applies a factor between 0 and 1 that uniformly scales the bids across all auctions.

We use anonymized and aggregated data from Facebook to explore the spatial structure of social networks in the New York metro area. We find that a substantial share of urban residents’ connections are to individuals who are located nearby.