A fundamental challenge in imitation learning is the \emph{covariate shift} problem. Existing methods to mitigate covariate shift often require additional expert interactions, access to environment dynamics, or complex adversarial training, which may not be practical in real-world applications. In this paper, we propose a simple yet effective method (DeCIL) to mitigate covariate shift by incorporating a denoising mechanism that enhances the contraction properties of the state transition mapping. Our approach involves training two neural networks: a dynamics model ( f ) that predicts the next state from the current state, and a joint state-action denoising policy network ( d ) that refines this state prediction via denoising and outputs the corresponding action. We provide theoretical analysis showing that the denoising network acts as a local contraction mapping, reducing the error propagation of the state transition and improving stability. Our method is straightforward to implement and can be easily integrated with existing imitation learning frameworks without requiring additional expert data or complex modifications to the training procedure. Empirical results demonstrate that our approach effectively improves success rate of various imitation learning tasks under noise perturbation.

Inspired by the ubiquitous use of differential equations to model continuous dynamics across diverse scientific and engineering domains, we propose a novel and intuitive approach to continuous sequence modeling. Our method interprets time-series data as \textit{discrete samples from an underlying continuous dynamical system}, and models its time evolution using Neural Stochastic Differential Equation (Neural SDE), where both the flow (drift) and diffusion terms are parameterized by neural networks. We derive a principled maximum likelihood objective and a \textit{simulation-free} scheme for efficient training of our Neural SDE model. We demonstrate the versatility of our approach through experiments on sequence modeling tasks across both embodied and generative AI. Notably, to the best of our knowledge, this is the first work to show that SDE-based continuous-time modeling also excels in such complex scenarios, and we hope that our work opens up new avenues for research of SDE models in high-dimensional and temporally intricate domains.

This paper presents an algorithmic framework for learning robust policies in asymmetric imperfect-information games, where the joint reward could depend on the uncertain opponent type (a private information known only to the opponent itself and its ally). In order to maximize the reward, the protagonist agent has to infer the opponent type through agent modeling. We use multiagent reinforcement learning (MARL) to learn opponent models through self-play, which captures the full strategy interaction and reasoning between agents. However, agent policies learned from self-play can suffer from mutual overfitting. Ensemble training methods can be used to improve the robustness of agent policy against different opponents, but it also significantly increases the computational overhead. In order to achieve a good trade-off between the robustness of the learned policy and the computation complexity, we propose to train a separate opponent policy against the protagonist agent for evaluation purposes. The reward achieved by this opponent is a noisy measure of the robustness of the protagonist agent policy due to the intrinsic stochastic nature of a reinforcement learner. To handle this stochasticity, we apply a stochastic optimization scheme to dynamically update the opponent ensemble to optimize an objective function that strikes a balance between robustness and computation complexity. We empirically show that, under the same limited computational budget, the proposed method results in more robust policy learning than standard ensemble training.

We pose an active perception problem where an autonomous agent actively interacts with a second agent with potentially adversarial behaviors. Given the uncertainty in the intent of the other agent, the objective is to collect further evidence to help discriminate potential threats. The main technical challenges are the partial observability of the agent intent, the adversary modeling, and the corresponding uncertainty modeling. Note that an adversary agent may act to mislead the autonomous agent by using a deceptive strategy that is learned from past experiences. We propose an approach that combines belief space planning, generative adversary modeling, and maximum entropy reinforcement learning to obtain a stochastic belief space policy. By accounting for various adversarial behaviors in the simulation framework and minimizing the predictability of the autonomous agent's action, the resulting policy is more robust to unmodeled adversarial strategies. This improved robustness is empirically shown against an adversary that adapts to and exploits the autonomous agent's policy when compared with a standard Chance-Constraint Partially Observable Markov Decision Process robust approach.

This project report compares some known GAN and VAE models proposed prior to 2017. There has been significant progress after we finished this report. We upload this report as an introduction to generative models and provide some personal interpretations supported by empirical evidence. Both generative adversarial network models and variational autoencoders have been widely used to approximate probability distributions of data sets. Although they both use parametrized distributions to approximate the underlying data distribution, whose exact inference is intractable, their behaviors are very different. We summarize our experiment results that compare these two categories of models in terms of fidelity and mode collapse. We provide a hypothesis to explain their different behaviors and propose a new model based on this hypothesis. We further tested our proposed model on MNIST dataset and CelebA dataset.