Offline reinforcement learning (RL) methods strike a balance between exploration and exploitation by conservative value estimation -- penalizing values of unseen states and actions. Model-free methods penalize values at all unseen actions, while model-based methods are able to further exploit unseen states via model rollouts. However, such methods are handicapped in their ability to find unseen states far away from the available offline data due to two factors -- (a) very short rollout horizons in models due to cascading model errors, and (b) model rollouts originating solely from states observed in offline data. We relax the second assumption and present a novel unseen state augmentation strategy to allow exploitation of unseen states where the learned model and value estimates generalize. Our strategy finds unseen states by value-informed perturbations of seen states followed by filtering out states with epistemic uncertainty estimates too high (high error) or too low (too similar to seen data). We observe improved performance in several offline RL tasks and find that our augmentation strategy consistently leads to overall lower average dataset Q-value estimates i.e. more conservative Q-value estimates than a baseline.
相關內容
Despite the promising progress in multi-modal tasks, current large multi-modal models (LMMs) are prone to hallucinating inconsistent descriptions with respect to the associated image and human instructions. This paper addresses this issue by introducing the first large and diverse visual instruction tuning dataset, named Large-scale Robust Visual (LRV)-Instruction. Our dataset comprises 400k visual instructions generated by GPT4, covering 16 vision-and-language tasks with open-ended instructions and answers. Unlike existing studies that primarily focus on positive instruction samples, we design LRV-Instruction to include both positive and negative instructions for more robust visual instruction tuning. Our negative instructions are designed at three semantic levels: (i) Nonexistent Object Manipulation, (ii) Existent Object Manipulation and (iii) Knowledge Manipulation. To efficiently measure the hallucination generated by LMMs, we propose GPT4-Assisted Visual Instruction Evaluation (GAVIE), a stable approach to evaluate visual instruction tuning like human experts. GAVIE does not require human-annotated groundtruth answers and can adapt to diverse instruction formats. We conduct comprehensive experiments to investigate the hallucination of LMMs. Our results demonstrate existing LMMs exhibit significant hallucinations when presented with our negative instructions, particularly Existent Object and Knowledge Manipulation instructions. Moreover, we successfully mitigate hallucination by finetuning MiniGPT4 and mPLUG-Owl on LRV-Instruction while improving performance on several public datasets compared to state-of-the-art methods. Additionally, we observed that a balanced ratio of positive and negative instances in the training data leads to a more robust model.
Pairwise comparison of graphs is key to many applications in Machine learning ranging from clustering, kernel-based classification/regression and more recently supervised graph prediction. Distances between graphs usually rely on informative representations of these structured objects such as bag of substructures or other graph embeddings. A recently popular solution consists in representing graphs as metric measure spaces, allowing to successfully leverage Optimal Transport, which provides meaningful distances allowing to compare them: the Gromov-Wasserstein distances. However, this family of distances overlooks edge attributes, which are essential for many structured objects. In this work, we introduce an extension of Gromov-Wasserstein distance for comparing graphs whose both nodes and edges have features. We propose novel algorithms for distance and barycenter computation. We empirically show the effectiveness of the novel distance in learning tasks where graphs occur in either input space or output space, such as classification and graph prediction.
Existing approaches defend against backdoor attacks in federated learning (FL) mainly through a) mitigating the impact of infected models, or b) excluding infected models. The former negatively impacts model accuracy, while the latter usually relies on globally clear boundaries between benign and infected model updates. However, model updates are easy to be mixed and scattered throughout in reality due to the diverse distributions of local data. This work focuses on excluding infected models in FL. Unlike previous perspectives from a global view, we propose Snowball, a novel anti-backdoor FL framework through bidirectional elections from an individual perspective inspired by one principle deduced by us and two principles in FL and deep learning. It is characterized by a) bottom-up election, where each candidate model update votes to several peer ones such that a few model updates are elected as selectees for aggregation; and b) top-down election, where selectees progressively enlarge themselves through picking up from the candidates. We compare Snowball with state-of-the-art defenses to backdoor attacks in FL on five real-world datasets, demonstrating its superior resistance to backdoor attacks and slight impact on the accuracy of the global model.
Product Lines (PL) have proved an effective approach to reuse-based systems development. Several modeling languages were proposed so far to specify PL. Although they can be very different, these languages show two common features: they emphasize (a) variability, and (b) the specification of constraints to define acceptable configurations. It is now widely acknowledged that configuring a product can be considered as a constraint satisfaction problem. It is thus natural to consider constraint programming as a first choice candidate to specify constraints on PL. For instance, the different constraints that can be specified using the FODA language can easily be expressed using boolean constraints, which enables automated calculation and configuration using a SAT solver. But constraint programming proposes other domains than the boolean domain: for instance integers, real, or sets. The integer domain was, for instance, proposed by Benavides to specify constraints on feature attributes. This paper proposes to further explore the use of integer constraint programming to specify PL constraints. The approach was implemented in a prototype tool. Its use in a real case showed that constraint programming encompasses different PL modeling languages (such as FORE, OVM, or else), and allows specifying complex constraints that are difficult to specify with these languages.
Rapidly developing machine learning methods has stimulated research interest in computationally reconstructing differential equations (DEs) from observational data which may provide additional insight into underlying causative mechanisms. In this paper, we propose a novel neural-ODE based method that uses spectral expansions in space to learn spatiotemporal DEs. The major advantage of our spectral neural DE learning approach is that it does not rely on spatial discretization, thus allowing the target spatiotemporal equations to contain long range, nonlocal spatial interactions that act on unbounded spatial domains. Our spectral approach is shown to be as accurate as some of the latest machine learning approaches for learning PDEs operating on bounded domains. By developing a spectral framework for learning both PDEs and integro-differential equations, we extend machine learning methods to apply to unbounded DEs and a larger class of problems.
We present a new algorithm based on posterior sampling for learning in constrained Markov decision processes (CMDP) in the infinite-horizon undiscounted setting. The algorithm achieves near-optimal regret bounds while being advantageous empirically compared to the existing algorithms. Our main theoretical result is a Bayesian regret bound for each cost component of \tilde{O} (HS \sqrt{AT}) for any communicating CMDP with S states, A actions, and bound on the hitting time H. This regret bound matches the lower bound in order of time horizon T and is the best-known regret bound for communicating CMDPs in the infinite-horizon undiscounted setting. Empirical results show that, despite its simplicity, our posterior sampling algorithm outperforms the existing algorithms for constrained reinforcement learning.
Reinforcement learning (RL) for bipedal locomotion has recently demonstrated robust gaits over moderate terrains using only proprioceptive sensing. However, such blind controllers will fail in environments where robots must anticipate and adapt to local terrain, which requires visual perception. In this paper, we propose a fully-learned system that allows bipedal robots to react to local terrain while maintaining commanded travel speed and direction. Our approach first trains a controller in simulation using a heightmap expressed in the robot's local frame. Next, data is collected in simulation to train a heightmap predictor, whose input is the history of depth images and robot states. We demonstrate that with appropriate domain randomization, this approach allows for successful sim-to-real transfer with no explicit pose estimation and no fine-tuning using real-world data. To the best of our knowledge, this is the first example of sim-to-real learning for vision-based bipedal locomotion over challenging terrains.
Pitch estimation is an essential step of many speech processing algorithms, including speech coding, synthesis, and enhancement. Recently, pitch estimators based on deep neural networks (DNNs) have have been outperforming well-established DSP-based techniques. Unfortunately, these new estimators can be impractical to deploy in real-time systems, both because of their relatively high complexity, and the fact that some require significant lookahead. We show that a hybrid estimator using a small deep neural network (DNN) with traditional DSP-based features can match or exceed the performance of pure DNN-based models, with a complexity and algorithmic delay comparable to traditional DSP-based algorithms. We further demonstrate that this hybrid approach can provide benefits for a neural vocoding task.
Recent contrastive representation learning methods rely on estimating mutual information (MI) between multiple views of an underlying context. E.g., we can derive multiple views of a given image by applying data augmentation, or we can split a sequence into views comprising the past and future of some step in the sequence. Contrastive lower bounds on MI are easy to optimize, but have a strong underestimation bias when estimating large amounts of MI. We propose decomposing the full MI estimation problem into a sum of smaller estimation problems by splitting one of the views into progressively more informed subviews and by applying the chain rule on MI between the decomposed views. This expression contains a sum of unconditional and conditional MI terms, each measuring modest chunks of the total MI, which facilitates approximation via contrastive bounds. To maximize the sum, we formulate a contrastive lower bound on the conditional MI which can be approximated efficiently. We refer to our general approach as Decomposed Estimation of Mutual Information (DEMI). We show that DEMI can capture a larger amount of MI than standard non-decomposed contrastive bounds in a synthetic setting, and learns better representations in a vision domain and for dialogue generation.
We propose a new method for event extraction (EE) task based on an imitation learning framework, specifically, inverse reinforcement learning (IRL) via generative adversarial network (GAN). The GAN estimates proper rewards according to the difference between the actions committed by the expert (or ground truth) and the agent among complicated states in the environment. EE task benefits from these dynamic rewards because instances and labels yield to various extents of difficulty and the gains are expected to be diverse -- e.g., an ambiguous but correctly detected trigger or argument should receive high gains -- while the traditional RL models usually neglect such differences and pay equal attention on all instances. Moreover, our experiments also demonstrate that the proposed framework outperforms state-of-the-art methods, without explicit feature engineering.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191