Beyond Flatland: A Geometric Take on Matching Methods for Treatment Effect Estimation & Efficient World Models with Context-Aware Tokenization

Add to your list(s) Download to your calendar using vCal

• Melanie Fernandez Pradier, Senior Researcher, MSRC & Vincent Micheli, University of Geneva
• Wednesday 11 September 2024, 17:30-19:00
• Microsoft Research Ltd, 198 Cambridge Science Park, Milton Road, CB4 0AB.

If you have a question about this talk, please contact Microsoft Research Cambridge Talks Admins.

Please note, this event may be recorded. Microsoft will own the copyright of any recording and reserves the right to distribute it as required

THESE TALKS WILL BE PRESENTED AT 198 CAMBRIDGE SCIENCE PARK NOT 21 STATION ROAD Speaker 1: Abstract : Matching is a popular approach in causal inference to estimate treatment effects by pairing treated and control units that are most similar in terms of their covariate information. However, classic matching methods completely ignore the geometry of the data manifold, which is crucial to define a meaningful distance for matching, and struggle when covariates are noisy and high-dimensional. In this work, we propose GeoMatching , a matching method to estimate treatment effects that takes into account the intrinsic data geometry induced by existing causal mechanisms among the confounding variables. First, we learn a low-dimensional, latent Riemannian manifold that accounts for uncertainty and geometry of the original input data. Second, we estimate treatment effects via matching in the latent space based on the learned latent Riemannian metric. We provide theoretical insights and empirical results in synthetic and real-world scenarios, demonstrating that GeoMatching yields more effective treatment effect estimators, even as we increase input dimensionality, in the presence of outliers, or in semi-supervised scenarios.

Speaker 2 Abstract: Scaling up deep Reinforcement Learning (RL) methods presents a significant challenge. Following developments in generative modelling, model-based RL positions itself as a strong contender. Recent advances in sequence modelling have led to effective transformer-based world models, albeit at the price of heavy computations due to the long sequences of tokens required to accurately simulate environments. In this work, we propose Δ-IRIS, a new agent with a world model architecture composed of a discrete autoencoder that encodes stochastic deltas between time steps and an autoregressive transformer that predicts future deltas by summarizing the current state of the world with continuous tokens. In the Crafter benchmark, Δ-IRIS sets a new state of the art at multiple frame budgets, while being an order of magnitude faster to train than previous attention-based approaches.

Please register: https://forms.office.com/pages/responsepage.aspx?id=v4j5cvGGr0GRqy180BHbR6ALaDloVjxDrC2QU3AErn5URURUV1NTQzhKS1VYWktESlA5TUVWR1AzRS4u&route=shorturl

This talk is part of the AI+Pizza series.

This talk is included in these lists:

• AI+Pizza
• Microsoft Research Ltd, 198 Cambridge Science Park, Milton Road, CB4 0AB
• bld31

Note that ex-directory lists are not shown.