Publications

Investigating Multi-Task Pretraining and Generalization in Reinforcement Learning
Adrien Ali Taiga
Rishabh Agarwal
Jesse Farebrother
Aaron Courville
Bellemare Marc-Emmanuel
Google Brain
2023-04-30
ICLR (Accept: poster)
openreview.net
openreview.net
Latent Bottlenecked Attentive Neural Processes
Leo Feng
Hossein Hajimirsadeghi
Yoshua Bengio
Mohamed Osama Ahmed
Neural Processes (NPs) are popular methods in meta-learning that can estimate predictive uncertainty on target datapoints by conditioning on… (voir plus) a context dataset. Previous state-of-the-art method Transformer Neural Processes (TNPs) achieve strong performance but require quadratic computation with respect to the number of context datapoints, significantly limiting its scalability. Conversely, existing sub-quadratic NP variants perform significantly worse than that of TNPs. Tackling this issue, we propose Latent Bottlenecked Attentive Neural Processes (LBANPs), a new computationally efficient sub-quadratic NP variant, that has a querying computational complexity independent of the number of context datapoints. The model encodes the context dataset into a constant number of latent vectors on which self-attention is performed. When making predictions, the model retrieves higher-order information from the context dataset via multiple cross-attention mechanisms on the latent vectors. We empirically show that LBANPs achieve results competitive with the state-of-the-art on meta-regression, image completion, and contextual multi-armed bandits. We demonstrate that LBANPs can trade-off the computational cost and performance according to the number of latent vectors. Finally, we show LBANPs can scale beyond existing attention-based NP variants to larger dataset settings.
2023-04-30
International Conference on Learning Representations (Accept: poster)
doi.org
openreview.net
Latent State Marginalization as a Low-Cost Approach for Improving Exploration
Dinghuai Zhang
Aaron Courville
Yoshua Bengio
Qinqing Zheng
Amy Zhang
Ricky T. Q. Chen
While the maximum entropy (MaxEnt) reinforcement learning (RL) framework -- often touted for its exploration and robustness capabilities -- … (voir plus)is usually motivated from a probabilistic perspective, the use of deep probabilistic models has not gained much traction in practice due to their inherent complexity. In this work, we propose the adoption of latent variable policies within the MaxEnt framework, which we show can provably approximate any policy distribution, and additionally, naturally emerges under the use of world models with a latent belief state. We discuss why latent variable policies are difficult to train, how naive approaches can fail, then subsequently introduce a series of improvements centered around low-cost marginalization of the latent state, allowing us to make full use of the latent state at minimal additional cost. We instantiate our method under the actor-critic framework, marginalizing both the actor and critic. The resulting algorithm, referred to as Stochastic Marginal Actor-Critic (SMAC), is simple yet effective. We experimentally validate our method on continuous control tasks, showing that effective marginalization can lead to better exploration and more robust training. Our implementation is open sourced at https://github.com/zdhNarsil/Stochastic-Marginal-Actor-Critic.
2023-04-30
International Conference on Learning Representations (Accept: poster)
doi.org
openreview.net
Learning on Large-Scale Text-Attributed Graphs via Variational Inference
Jianan Zhao
Meng Qu
Chaozhuo Li
Hao Yan
Qian Liu
Rui Li
Xing Xie
Jian Tang
This paper studies learning on text-attributed graphs (TAGs), where each node is associated with a text description. An ideal solution for s… (voir plus)uch a problem would be integrating both the text and graph structure information with large language models and graph neural networks (GNNs). However, the problem becomes very challenging when graphs are large due to the high computational complexity brought by training large language models and GNNs together. In this paper, we propose an efficient and effective solution to learning on large text-attributed graphs by fusing graph structure and language learning with a variational Expectation-Maximization (EM) framework, called GLEM. Instead of simultaneously training large language models and GNNs on big graphs, GLEM proposes to alternatively update the two modules in the E-step and M-step. Such a procedure allows training the two modules separately while simultaneously allowing the two modules to interact and mutually enhance each other. Extensive experiments on multiple data sets demonstrate the efficiency and effectiveness of the proposed approach.
2023-04-30
International Conference on Learning Representations (Accept: notable-top-5%)
doi.org
openreview.net
Low-Complexity Sphere Decoding for Polar-Coded MIMO Systems
Huayi Zhou
Jian Zheng
Minhua Yang
Warren J. Gross
Xiaohu You
Chuan Zhang
For polar-coded MIMO systems, separate detection and decoding (SDD) is the traditional scheme. In SDD systems, sphere decoding (SD) is one o… (voir plus)f the competitive MIMO detection schemes. However, SD may not utilize the coding information sufficiently in SDD systems, causing an error-correction performance loss. The existed joint detection and decoding using breadth-first SD (BSD) improves the performance than SDD, whereas the limited search space still causes a performance loss. In this paper, we propose joint detection and decoding based on SD (SD JDD) for polar-coded MIMO systems to reach maximum likelihood (ML) bound. Subsequently, two approaches are further proposed to reduce the computational complexity. The first approach reduces the layers of the SD search tree by exploiting symbol synchro sets, which could accelerate the convergence of SD JDD. The second efficient approach performs multiple tree searches. A small initial radius of the sphere for the first search is assigned to reduce the search space. The ML optimality could be preserved by the following multiple tree searches with increasing radius. It is shown from the numerical results that the proposed JDD outperforms SDD by 3.1 dB at FER
2023-04-30
IEEE Transactions on Vehicular Technology (publié)
doi.org
MAPL: Parameter-Efficient Adaptation of Unimodal Pre-Trained Models for Vision-Language Few-Shot Prompting
Oscar Mañas
Pau Rodríguez
Saba Ahmadi
Aida Nematzadeh
Yash Goyal
Aishwarya Agrawal
Large pre-trained models have proved to be remarkable zero- and (prompt-based) few-shot learners in unimodal vision and language tasks. We p… (voir plus)ropose MAPL, a simple and parameter-efficient method that reuses frozen pre-trained unimodal models and leverages their strong generalization capabilities in multimodal vision-language (VL) settings. MAPL learns a lightweight mapping between the representation spaces of unimodal models using aligned image-text data, and can generalize to unseen VL tasks from just a few in-context examples. The small number of trainable parameters makes MAPL effective at low-data and in-domain learning. Moreover, MAPL's modularity enables easy extension to other pre-trained models. Extensive experiments on several visual question answering and image captioning benchmarks show that MAPL achieves superior or competitive performance compared to similar methods while training orders of magnitude fewer parameters. MAPL can be trained in just a few hours using modest computational resources and public datasets. We release our code and pre-trained model weights at https://github.com/mair-lab/mapl.
2023-04-30
Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics (publié)
doi.org
arxiv.org
MeshDiffusion: Score-Based Generative 3D Mesh Modeling
Zhen Liu
Yao Feng
Michael J. Black
Derek Nowrouzezahrai
Liam Paull
Weiyang Liu
We consider the task of generating realistic 3D shapes, which is useful for a variety of applications such as automatic scene generation and… (voir plus) physical simulation. Compared to other 3D representations like voxels and point clouds, meshes are more desirable in practice, because (1) they enable easy and arbitrary manipulation of shapes for relighting and simulation, and (2) they can fully leverage the power of modern graphics pipelines which are mostly optimized for meshes. Previous scalable methods for generating meshes typically rely on sub-optimal post-processing, and they tend to produce overly-smooth or noisy surfaces without fine-grained geometric details. To overcome these shortcomings, we take advantage of the graph structure of meshes and use a simple yet very effective generative modeling method to generate 3D meshes. Specifically, we represent meshes with deformable tetrahedral grids, and then train a diffusion model on this direct parametrization. We demonstrate the effectiveness of our model on multiple generative tasks.
2023-04-30
International Conference on Learning Representations (Accept: notable-top-25%)
doi.org
openreview.net
Molecular Geometry Pretraining with SE(3)-Invariant Denoising Distance Matching
Shengchao Liu
Hongyu Guo
Jian Tang
Molecular representation pretraining is critical in various applications for drug and material discovery due to the limited number of labele… (voir plus)d molecules, and most existing work focuses on pretraining on 2D molecular graphs. However, the power of pretraining on 3D geometric structures has been less explored. This is owing to the difficulty of finding a sufficient proxy task that can empower the pretraining to effectively extract essential features from the geometric structures. Motivated by the dynamic nature of 3D molecules, where the continuous motion of a molecule in the 3D Euclidean space forms a smooth potential energy surface, we propose GeoSSL, a 3D coordinate denoising pretraining framework to model such an energy landscape. Further by leveraging an SE(3)-invariant score matching method, we propose GeoSSL-DDM in which the coordinate denoising proxy task is effectively boiled down to denoising the pairwise atomic distances in a molecule. Our comprehensive experiments confirm the effectiveness and robustness of our proposed method.
2023-04-30
International Conference on Learning Representations (Accept: poster)
doi.org
openreview.net
Neural Networks Efficiently Learn Low-Dimensional Representations with SGD
Alireza Mousavi-Hosseini
Sejun Park
Manuela Girotti
Ioannis Mitliagkas
Murat A. Erdogdu
We study the problem of training a two-layer neural network (NN) of arbitrary width using stochastic gradient descent (SGD) where the input …
2023-04-30
International Conference on Learning Representations (Accept: notable-top-25%)
doi.org
openreview.net
OC-0290 Investigation of the feasibility of selenium-75 as a viable brachytherapy source
J. Reid
Jonathan Kalinowski
J. Munro
A. Armstrong
S. Enger
2023-04-30
Radiotherapy and Oncology (publié)
doi.org
PD-0334 Techniques to optimize auto-segmentation of small OARs in pediatric patients undergoing CSI
J. Tsui
M. Popovic
O. Ates
C. Hua
J. Schneider
S. Skamene
C. Freeman
S. Enger
2023-04-30
Radiotherapy and Oncology (publié)
doi.org
PD-0505 Monte Carlo simulated correction factors of a novel phantom for brachytherapy dosimetry audits
K. Chelminski
R. Abdulrahim
A. Dimitriadis
E. Granizo-Roman
Jonathan Kalinowski
S. Enger
G. Azangwe
J. Swamidas
2023-04-30
Radiotherapy and Oncology (publié)
doi.org