Imitation learning from demonstrations (ILD) aims to alleviate numerous shortcomings of reinforcement learning through the use of demonstrations. However, in most real-world applications, expert action guidance is absent, making the use of ILD impossible. Instead, we consider imitation learning from observations (ILO), where no expert actions are provided, making it a significantly more challenging problem to address. Existing methods often employ on-policy learning, which is known to be sample-costly. This paper presents SEILO, a novel sample-efficient on-policy algorithm for ILO, that combines standard adversarial imitation learning with inverse dynamics modeling. This approach enables the agent to receive feedback from both the adversarial procedure and a behavior cloning loss. We empirically demonstrate that our proposed algorithm requires fewer interactions with the environment to achieve expert performance compared to other state-of-the-art on-policy ILO and ILD methods.
相關內容
After completing the design and training phases, deploying a deep learning model onto specific hardware is essential before practical implementation. Targeted optimizations are necessary to enhance the model's performance by reducing inference latency. Auto-scheduling, an automated technique offering various optimization options, proves to be a viable solution for large-scale auto-deployment. However, the low-level code generated by auto-scheduling resembles hardware coding, potentially hindering human comprehension and impeding manual optimization efforts. In this ongoing study, we aim to develop an enhanced visualization that effectively addresses the extensive profiling metrics associated with auto-scheduling. This visualization will illuminate the intricate scheduling process, enabling further advancements in latency optimization through insights derived from the schedule.
Homomorphic encryption (HE) has found extensive utilization in federated learning (FL) systems, capitalizing on its dual advantages: (i) ensuring the confidentiality of shared models contributed by participating entities, and (ii) enabling algebraic operations directly on ciphertexts representing encrypted models. Particularly, the approximate fully homomorphic encryption (FHE) scheme, known as CKKS, has emerged as the de facto encryption scheme, notably supporting decimal numbers. While recent research predominantly focuses on enhancing CKKS's encryption rate and evaluation speed in the context of FL, the search operation has been relatively disregarded due to the tendency of some applications to discard intermediate encrypted models. Yet, emerging studies emphasize the importance of managing and searching intermediate models for specific applications like large-scale scientific computing, necessitating robust data provenance and auditing support. To address this, our paper introduces an innovative approach that efficiently searches for a target encrypted value, incurring only a logarithmic number of network interactions. The proposed method capitalizes on CKKS's additive and multiplicative properties on encrypted models, propagating equality comparisons between values through a balanced binary tree structure to ultimately reach a single aggregate. A comprehensive analysis of the proposed algorithm underscores its potential to significantly broaden FL's applicability and impact.
To use reinforcement learning from human feedback (RLHF) in practical applications, it is crucial to learn reward models from diverse sources of human feedback and to consider human factors involved in providing feedback of different types. However, the systematic study of learning from diverse types of feedback is held back by limited standardized tooling available to researchers. To bridge this gap, we propose RLHF-Blender, a configurable, interactive interface for learning from human feedback. RLHF-Blender provides a modular experimentation framework and implementation that enables researchers to systematically investigate the properties and qualities of human feedback for reward learning. The system facilitates the exploration of various feedback types, including demonstrations, rankings, comparisons, and natural language instructions, as well as studies considering the impact of human factors on their effectiveness. We discuss a set of concrete research opportunities enabled by RLHF-Blender. More information is available at //rlhfblender.info/.
Federated learning (FL) has been widely deployed to enable machine learning training on sensitive data across distributed devices. However, the decentralized learning paradigm and heterogeneity of FL further extend the attack surface for backdoor attacks. Existing FL attack and defense methodologies typically focus on the whole model. None of them recognizes the existence of backdoor-critical (BC) layers-a small subset of layers that dominate the model vulnerabilities. Attacking the BC layers achieves equivalent effects as attacking the whole model but at a far smaller chance of being detected by state-of-the-art (SOTA) defenses. This paper proposes a general in-situ approach that identifies and verifies BC layers from the perspective of attackers. Based on the identified BC layers, we carefully craft a new backdoor attack methodology that adaptively seeks a fundamental balance between attacking effects and stealthiness under various defense strategies. Extensive experiments show that our BC layer-aware backdoor attacks can successfully backdoor FL under seven SOTA defenses with only 10% malicious clients and outperform the latest backdoor attack methods.
Compositional zero-shot learning (CZSL) aims to recognize unseen compositions with prior knowledge of known primitives (attribute and object). Previous works for CZSL often suffer from grasping the contextuality between attribute and object, as well as the discriminability of visual features, and the long-tailed distribution of real-world compositional data. We propose a simple and scalable framework called Composition Transformer (CoT) to address these issues. CoT employs object and attribute experts in distinctive manners to generate representative embeddings, using the visual network hierarchically. The object expert extracts representative object embeddings from the final layer in a bottom-up manner, while the attribute expert makes attribute embeddings in a top-down manner with a proposed object-guided attention module that models contextuality explicitly. To remedy biased prediction caused by imbalanced data distribution, we develop a simple minority attribute augmentation (MAA) that synthesizes virtual samples by mixing two images and oversampling minority attribute classes. Our method achieves SoTA performance on several benchmarks, including MIT-States, C-GQA, and VAW-CZSL. We also demonstrate the effectiveness of CoT in improving visual discrimination and addressing the model bias from the imbalanced data distribution. The code is available at //github.com/HanjaeKim98/CoT.
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.
Supervised learning algorithms generally assume the availability of enough memory to store their data model during the training and test phases. However, in the Internet of Things, this assumption is unrealistic when data comes in the form of infinite data streams, or when learning algorithms are deployed on devices with reduced amounts of memory. In this paper, we adapt the online Mondrian forest classification algorithm to work with memory constraints on data streams. In particular, we design five out-of-memory strategies to update Mondrian trees with new data points when the memory limit is reached. Moreover, we design trimming mechanisms to make Mondrian trees more robust to concept drifts under memory constraints. We evaluate our algorithms on a variety of real and simulated datasets, and we conclude with recommendations on their use in different situations: the Extend Node strategy appears as the best out-of-memory strategy in all configurations, whereas different trimming mechanisms should be adopted depending on whether a concept drift is expected. All our methods are implemented in the OrpailleCC open-source library and are ready to be used on embedded systems and connected objects.
Recently, researchers have gradually realized that in some cases, the self-supervised pre-training on large-scale Internet data is better than that of high-quality/manually labeled data sets, and multimodal/large models are better than single or bimodal/small models. In this paper, we propose a robust audio representation learning method WavBriVL based on Bridging-Vision-and-Language (BriVL). WavBriVL projects audio, image and text into a shared embedded space, so that multi-modal applications can be realized. We demonstrate the qualitative evaluation of the image generated from WavBriVL as a shared embedded space, with the main purposes of this paper:(1) Learning the correlation between audio and image;(2) Explore a new way of image generation, that is, use audio to generate pictures. Experimental results show that this method can effectively generate appropriate images from audio.
This paper aims to mitigate straggler effects in synchronous distributed learning for multi-agent reinforcement learning (MARL) problems. Stragglers arise frequently in a distributed learning system, due to the existence of various system disturbances such as slow-downs or failures of compute nodes and communication bottlenecks. To resolve this issue, we propose a coded distributed learning framework, which speeds up the training of MARL algorithms in the presence of stragglers, while maintaining the same accuracy as the centralized approach. As an illustration, a coded distributed version of the multi-agent deep deterministic policy gradient(MADDPG) algorithm is developed and evaluated. Different coding schemes, including maximum distance separable (MDS)code, random sparse code, replication-based code, and regular low density parity check (LDPC) code are also investigated. Simulations in several multi-robot problems demonstrate the promising performance of the proposed framework.
Deep learning methods are achieving ever-increasing performance on many artificial intelligence tasks. A major limitation of deep models is that they are not amenable to interpretability. This limitation can be circumvented by developing post hoc techniques to explain the predictions, giving rise to the area of explainability. Recently, explainability of deep models on images and texts has achieved significant progress. In the area of graph data, graph neural networks (GNNs) and their explainability are experiencing rapid developments. However, there is neither a unified treatment of GNN explainability methods, nor a standard benchmark and testbed for evaluations. In this survey, we provide a unified and taxonomic view of current GNN explainability methods. Our unified and taxonomic treatments of this subject shed lights on the commonalities and differences of existing methods and set the stage for further methodological developments. To facilitate evaluations, we generate a set of benchmark graph datasets specifically for GNN explainability. We summarize current datasets and metrics for evaluating GNN explainability. Altogether, this work provides a unified methodological treatment of GNN explainability and a standardized testbed for evaluations.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191