The latent spaces of typical GAN models often have semantically meaningful directions. Moving in these directions corresponds to human-interpretable image transformations, such as zooming or recoloring, enabling a more controllable generation process. However, the discovery of such directions is currently performed in a supervised manner, requiring human labels, pretrained models, or some form of self-supervision. These requirements can severely limit a range of directions existing approaches can discover. In this paper, we introduce an unsupervised method to identify interpretable directions in the latent space of a pretrained GAN model. By a simple model-agnostic procedure, we find directions corresponding to sensible semantic manipulations without any form of (self-)supervision. Furthermore, we reveal several non-trivial findings, which would be difficult to obtain by existing methods, e.g., a direction corresponding to background removal. As an immediate practical benefit of our work, we show how to exploit this finding to achieve a new state-of-the-art for the problem of saliency detection.

Large-scale nonconvex optimization problems are ubiquitous in modern machine learning, and among practitioners interested in solving them, Stochastic Gradient Descent (SGD) reigns supreme. We revisit the analysis of SGD in the nonconvex setting and propose a new variant of the recently introduced expected smoothness assumption which governs the behaviour of the second moment of the stochastic gradient. We show that our assumption is both more general and more reasonable than assumptions made in all prior work. Moreover, our results yield the optimal $\mathcal{O}(\varepsilon^{-4})$ rate for finding a stationary point of nonconvex smooth functions, and recover the optimal $\mathcal{O}(\varepsilon^{-1})$ rate for finding a global solution if the Polyak-{\L}ojasiewicz condition is satisfied. We compare against convergence rates under convexity and prove a theorem on the convergence of SGD under Quadratic Functional Growth and convexity, which might be of independent interest. Moreover, we perform our analysis in a framework which allows for a detailed study of the effects of a wide array of sampling strategies and minibatch sizes for finite-sum optimization problems. We corroborate our theoretical results with experiments on real and synthetic data.

Random utility models (RUMs) are a popular and well-established framework for studying behavioral choices by individuals and groups Thurstone [1927]. In a RUM with finite alternatives or items, a distribution on the preferred alternative(s) is assumed to arise from a random utility drawn from a distribution for each item, followed by rank ordering the items according to their utilities. Perhaps the most widely known RUM is the Plackett-Luce or multinomial logit model Plackett [1975], Luce [2012] which results when each item's utility is sampled from an additive model with a Gumbel-distributed perturbation. It is unique in the sense of enjoying the property of independence of irrelevant attributes (IIA), which is often key in permitting efficient inference of Plackett-Luce models from data Khetan and Oh [2016]. Other well-known RUMs include the probit model Bliss [1934] featuring random Gaussian perturbations to the intrinsic utilities, mixed logit, nested logit, etc. A long line of work in statistics and machine learning focuses on estimating RUM properties from observed data Soufiani et al. [2014], Zhao et al. [2018], Soufiani et al. [2013].

Attribution methods that explains the behaviour of machine learning models, e.g. Convolutional Neural Networks (CNNs), have developed into many different forms, motivated by desirable distinct, though related, criteria. Following the diversity of attribution methods, evaluation tools are in need to answer: which method is better for what purpose and why? This paper introduces a new way to decompose the evaluation for attribution methods into two criteria: ordering and proportionality. We argue that existing evaluations follow an ordering criteria roughly corresponding to either the logical concept of necessity or sufficiency. The paper further demonstrates a notion of Proportionality for Necessity and Sufficiency, a quantitative evaluation to compare existing attribution methods, as a refinement to the ordering criteria. Evaluating the performance of existing attribution methods on explaining the CNN for image classification, we conclude that some attribution methods are better in the necessity analysis and the others are better in the sufficiency analysis, but no method is always the winner on both sides.

Gradient-based meta-learners such as Model-Agnostic Meta-Learning (MAML) have shown strong few-shot performance in supervised and reinforcement learning settings. However, specifically in the case of meta-reinforcement learning (meta-RL), we can show that gradient-based meta-learners are sensitive to task distributions. With the wrong curriculum, agents suffer the effects of meta-overfitting, shallow adaptation, and adaptation instability. In this work, we begin by highlighting intriguing failure cases of gradient-based meta-RL and show that task distributions can wildly affect algorithmic outputs, stability, and performance. To address this problem, we leverage insights from recent literature on domain randomization and propose meta Active Domain Randomization (meta-ADR), which learns a curriculum of tasks for gradient-based meta-RL in a similar as ADR does for sim2real transfer. We show that this approach induces more stable policies on a variety of simulated locomotion and navigation tasks. We assess in- and out-of-distribution generalization and find that the learned task distributions, even in an unstructured task space, greatly improve the adaptation performance of MAML. Finally, we motivate the need for better benchmarking in meta-RL that prioritizes \textit{generalization} over single-task adaption performance.

Since its inception, Facebook has become an integral part of the online social community. People rely on Facebook to make connections with others and build communities. As a result, it is paramount to protect the integrity of such a rapidly growing network in a fast and scalable manner. In this paper, we present our efforts to protect various social media entities at Facebook from people who try to abuse our platform. We present a novel Temporal Interaction EmbeddingS (TIES) model that is designed to capture rogue social interactions and flag them for further suitable actions. TIES is a supervised, deep learning, production ready model at Facebook-scale networks. Prior works on integrity problems are mostly focused on capturing either only static or certain dynamic features of social entities. In contrast, TIES can capture both these variant behaviors in a unified model owing to the recent strides made in the domains of graph embedding and deep sequential pattern learning. To show the real-world impact of TIES, we present a few applications especially for preventing spread of misinformation, fake account detection, and reducing ads payment risks in order to enhance the platform's integrity.

Goal-directed Reinforcement Learning (RL) traditionally considers an agent interacting with an environment, prescribing a real-valued reward to an agent proportional to the completion of some goal. Goal-directed RL has seen large gains in sample efficiency, due to the ease of reusing or generating new experience by proposing goals. In this work, we build on the framework of self-play, allowing an agent to interact with itself in order to make progress on some unknown task. We use Active Domain Randomization and self-play to create a novel, coupled environment-goal curriculum, where agents learn through progressively more difficult tasks and environment variations. Our method, Self-Supervised Active Domain Randomization (SS-ADR), generates a growing curriculum, encouraging the agent to try tasks that are just outside of its current capabilities, while building a domain-randomization curriculum that enables state-of-the-art results on various sim2real transfer tasks. Our results show that a curriculum of co-evolving the environment difficulty along with the difficulty of goals set in each environment provides practical benefits in the goal-directed tasks tested.

Tree-based Long short term memory (LSTM) network has become state-of-the-art for modeling the meaning of language texts as they can effectively exploit the grammatical syntax and thereby non-linear dependencies among words of the sentence. However, most of these models cannot recognize the difference in meaning caused by a change in semantic roles of words or phrases because they do not acknowledge the type of grammatical relations, also known as typed dependencies, in sentence structure. This paper proposes an enhanced LSTM architecture, called relation gated LSTM, which can model the relationship between two inputs of a sequence using a control input. We also introduce a Tree-LSTM model called Typed Dependency Tree-LSTM that uses the sentence dependency parse structure as well as the dependency type to embed sentence meaning into a dense vector. The proposed model outperformed its type-unaware counterpart in two typical NLP tasks - Semantic Relatedness Scoring and Sentiment Analysis, in a lesser number of training epochs. The results were comparable or competitive with other state-of-the-art models. Qualitative analysis showed that changes in the voice of sentences had little effect on the model's predicted scores, while changes in nominal (noun) words had a more significant impact. The model recognized subtle semantic relationships in sentence pairs. The magnitudes of learned typed dependencies embeddings were also in agreement with human intuitions. The research findings imply the significance of grammatical relations in sentence modeling. The proposed models would serve as a base for future researches in this direction.

Integrated Information Theory is one of the leading models of consciousness. It aims to describe both the quality and quantity of the conscious experience of a physical system, such as the brain, in a particular state. In this contribution, we propound the mathematical structure of the theory, separating the essentials from auxiliary formal tools. We provide a definition of a generalized IIT which has IIT 3.0 of Tononi et. al., as well as the Quantum IIT introduced by Zanardi et. al. as special cases. This provides an axiomatic definition of the theory which may serve as the starting point for future formal investigations and as an introduction suitable for researchers with a formal background.

Accurate and timely metro passenger flow forecasting is critical for the successful deployment of intelligent transportation systems. However, it is quite challenging to propose an efficient and robust forecasting approach due to the inherent randomness and variations of metro passenger flow. In this study, we present a novel adaptive ensemble (AdaEnsemble) learning approach to accurately forecast the volume of metro passenger flows, and it combines the complementary advantages of variational mode decomposition (VMD), seasonal autoregressive integrated moving averaging (SARIMA), multilayer perceptron network (MLP) and long short-term memory (LSTM) network. The AdaEnsemble learning approach consists of three important stages. The first stage applies VMD to decompose the metro passenger flows data into periodic component, deterministic component and volatility component. Then we employ SARIMA model to forecast the periodic component, LSTM network to learn and forecast deterministic component and MLP network to forecast volatility component. In the last stage, the diverse forecasted components are reconstructed by another MLP network. The empirical results show that our proposed AdaEnsemble learning approach not only has the best forecasting performance compared with the state-of-the-art models but also appears to be the most promising and robust based on the historical passenger flow data in Shenzhen subway system and several standard evaluation measures.