Deep Q Network (DQN) firstly kicked the door of deep reinforcement learning (DRL) via combining deep learning (DL) with reinforcement learning (RL), which has noticed that the distribution of the acquired data would change during the training process. DQN found this property might cause instability for training, so it proposed effective methods to handle the downside of the property. Instead of focusing on the unfavourable aspects, we find it critical for RL to ease the gap between the estimated data distribution and the ground truth data distribution while supervised learning (SL) fails to do so. From this new perspective, we extend the basic paradigm of RL called the Generalized Policy Iteration (GPI) into a more generalized version, which is called the Generalized Data Distribution Iteration (GDI). We see massive RL algorithms and techniques can be unified into the GDI paradigm, which can be considered as one of the special cases of GDI. We provide theoretical proof of why GDI is better than GPI and how it works. Several practical algorithms based on GDI have been proposed to verify the effectiveness and extensiveness of it. Empirical experiments prove our state-of-the-art (SOTA) performance on Arcade Learning Environment (ALE), wherein our algorithm has achieved 9620.98% mean human normalized score (HNS), 1146.39% median HNS and 22 human world record breakthroughs (HWRB) using only 200M training frames. Our work aims to lead the RL research to step into the journey of conquering the human world records and seek real superhuman agents on both performance and efficiency.
Tsinghua University（清华大学）；University of Cambridge（剑桥大学）；Eindhoven University of Technology（埃因霍温理工大学）；King’s College London（伦敦国王学院）；Tencent Robotics X（腾讯机器人实验室）
Solving goal-conditioned tasks with sparse rewards using self-supervised learning is promising because of its simplicity and stability over current reinforcement learning (RL) algorithms. A recent work, called Goal-Conditioned Supervised Learning (GCSL), provides a new learning framework by iteratively relabeling and imitating self-generated experiences. In this paper, we revisit the theoretical property of GCSL -- optimizing a lower bound of the goal reaching objective, and extend GCSL as a novel offline goal-conditioned RL algorithm. The proposed method is named Weighted GCSL (WGCSL), in whi…
University of Amsterdam（阿姆斯特丹大学）；Microsoft Research（微软研究院）
Dialogue policy learning for task-oriented dialogue systems has enjoyed great progress recently mostly through employing reinforcement learning methods. However, these approaches have become very sophisticated. It is time to re-evaluate it. Are we really making progress developing dialogue agents only based on reinforcement learning? We demonstrate how (1)~traditional supervised learning together with (2)~a simulator-free adversarial learning method can be used to achieve performance comparable to state-of-the-art RL-based methods. First, we introduce a simple dialogue action decoder to predic…
We are at the cusp of a transition from "learning from data" to "learning what data to learn from" as a central focus of artificial intelligence (AI) research. While the first-order learning problem is not completely solved, large models under unified architectures, such as transformers, have shifted the learning bottleneck from how to effectively train our models to how to effectively acquire and use task-relevant data. This problem, which we frame as exploration, is a universal aspect of learning in open-ended domains, such as the real world. Although the study of exploration in AI is largel…
Seoul National University（首尔国立大学）；Neural Processing Research Center（神经处理研究中心）；DeepMetrics
Our work focuses on training RL agents on multiple visually diverse environments to improve observational generalization performance. In prior methods, policy and value networks are separately optimized using a disjoint network architecture to avoid interference and obtain a more accurate value function. We identify that a value network in the multi-environment setting is more challenging to optimize and prone to memorizing the training data than in the conventional single-environment setting. In addition, we find that appropriate regularization on the value network is necessary to improve bot…
Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening
Shanghai Jiao Tong University（上海交通大学）；Shanghai AI Laboratory（上海人工智能实验室）；Westlake University（西湖大学）；Nanjing University（南京大学）；Tsinghua University（清华大学）；The Chinese University of Hong Kong（香港中文大学）；Nanyang Technological University（南洋理工大学）
Semi-supervised learning methods can train high-accuracy machine learning models with a fraction of the labeled training samples required for traditional supervised learning. Such methods do not typically involve close review of the unlabeled training samples, making them tempting targets for data poisoning attacks. In this paper we investigate the vulnerabilities of semi-supervised learning methods to backdoor data poisoning attacks on the unlabeled samples. We show that simple poisoning attacks that influence the distribution of the poisoned samples' predicted labels are highly effective - a…
National Key Laboratory for Novel Software Technology（计算机软件新技术国家重点实验室）；Nanjing University（南京大学）
Self-supervised learning (SSL), in particular contrastive learning, has made great progress in recent years. However, a common theme in these methods is that they inherit the learning paradigm from the supervised deep learning scenario. Current SSL methods are often pretrained for many epochs on large-scale datasets using high resolution images, which brings heavy computational cost and lacks flexibility. In this paper, we demonstrate that the learning paradigm for SSL should be different from supervised learning and the information encoded by the contrastive loss is expected to be much less t…
Mila（米拉人工智能研究所）；Université de Montréal（蒙特利尔大学）；University of Osnabrück（奥斯纳布吕克大学）；Institute of Cognitive Science（认知科学研究所）
The renaissance of artificial neural networks was catalysed by the success of classification models, tagged by the community with the broader term supervised learning. The extraordinary results gave rise to a hype loaded with ambitious promises and overstatements. Soon the community realised that the success owed much to the availability of thousands of labelled examples and supervised learning went, for many, from glory to shame: Some criticised deep learning as a whole and others proclaimed that the way forward had to be alternatives to supervised learning: predictive, unsupervised, semi-sup…
We argue that the present setting of semisupervised learning on graphs may result in unfair comparisons, due to its potential risk of over-tuning hyper-parameters for models. In this paper, we highlight the significant influence of tuning hyper-parameters, which leverages the label information in the validation set to improve the performance. To explore the limit of over-tuning hyperparameters, we propose ValidUtil, an approach to fully utilize the label information in the validation set through an extra group of hyper-parameters. With ValidUtil, even GCN can easily get high accuracy of 85.8% …
Graph Contrastive Learning (GCL) has proven highly effective in promoting the performance of Semi-Supervised Node Classification (SSNC). However, existing GCL methods are generally transferred from other fields like CV or NLP, whose underlying working mechanism remains under-explored. In this work, we first deeply probe the working mechanism of GCL in SSNC, and find that the promotion brought by GCL is severely unevenly distributed: the improvement mainly comes from subgraphs with less annotated information, which is fundamentally different from contrastive learning in other fields. However, e…
The pretrain-finetune paradigm has shown outstanding performance on many applications of deep learning, where a model is pre-trained on a upstream large dataset (e.g. ImageNet), and is then fine-tuned to different downstream tasks. Though for most cases, the pre-training stage is conducted based on supervised methods, recent works on self-supervised pre-training have shown powerful transferability and even outperform supervised pre-training on multiple downstream tasks. It thus remains an open question how to better generalize supervised pre-training model to downstream tasks. In this paper, w…
Recent works have shown that self-supervised learning can achieve remarkable robustness when integrated with adversarial training (AT). However, the robustness gap between supervised AT (sup-AT) and self-supervised AT (self-AT) remains significant. Motivated by this observation, we revisit existing self-AT methods and discover an inherent dilemma that affects self-AT robustness: either strong or weak data augmentations are harmful to self-AT, and a medium strength is insufficient to bridge the gap. To resolve this dilemma, we propose a simple remedy named DYNACL (Dynamic Adversarial Contrastiv…
Graph contrastive learning (GCL) alleviates the heavy reliance on label information for graph representation learning (GRL) via self-supervised learning schemes. The core idea is to learn by maximising mutual information for similar instances, which requires similarity computation between two node instances. However, GCL is inefficient in both time and memory consumption. In addition, GCL normally requires a large number of training epochs to be well-trained on large-scale datasets. Inspired by an observation of a technical defect (i.e., inappropriate usage of Sigmoid function) commonly used i…
Google Research（谷歌研究院）；MIT CSAIL（麻省理工学院计算机科学与人工智能实验室）
The focus of recent meta-learning research has been on the development of learning algorithms that can quickly adapt to test time tasks with limited data and low computational cost. Few-shot learning is widely used as one of the standard benchmarks in meta-learning. In this work, we show that a simple baseline: learning a supervised or self-supervised representation on the meta-training set, followed by training a linear classifier on top of this representation, outperforms state-of-the-art few-shot learning methods. An additional boost can be achieved through the use of self-distillation. Thi…
Rethinking Independent Cross-Entropy Loss For Graph-Structured Data
School of Artificial Intelligence, Jilin University（吉林大学人工智能学院）；Institute for Medical Engineering Science, Massachusetts Institute of Technology（麻省理工学院医学工程科学研究所）；School of Computing, University of Georgia（佐治亚大学计算学院）；College of Computer Science and Technology, Jilin University（吉林大学计算机科学与技术学院）
Massachusetts Institute of Technology（麻省理工学院）；EECS（电气工程与计算机科学系（麻省理工学院））
Real-world data often exhibits long-tailed distributions with heavy class imbalance, posing great challenges for deep recognition models. We identify a persisting dilemma on the value of labels in the context of imbalanced learning: on the one hand, supervision from labels typically leads to better results than its unsupervised counterparts; on the other hand, heavily imbalanced data naturally incurs "label bias" in the classifier, where the decision boundary can be drastically altered by the majority classes. In this work, we systematically investigate these two facets of labels. We demonstra…
Labels Matter More Than Models: Rethinking the Unsupervised Paradigm in Time Series Anomaly Detection
School of Future Technology, South China University of Technology（华南理工大学未来技术学院）；Pengcheng Laboratory（鹏城实验室）；School of Computer Science and Engineering, South China University of Technology（华南理工大学计算机科学与工程学院）
Rethinking Knee Osteoarthritis Severity Grading: A Few Shot Self-Supervised Contrastive Learning Approach
Science Foundation Ireland Centre for Research Training in Machine Learning（爱尔兰科学基金会机器学习研究培训中心）；School of Medicine（医学院）；School of Computer Science, University College Dublin（都柏林大学学院计算机科学学院）；Insight Centre for Data Analytics, University College Dublin（都柏林大学学院洞察数据分析中心）
Inception Institute of Artificial Intelligence（初始人工智能研究院）；University of Macau（澳门大学）；Terminus Group（特斯联集团）
The pioneering method for unsupervised meta-learning, CACTUs, is a clustering-based approach with pseudo-labeling. This approach is model-agnostic and can be combined with supervised algorithms to learn from unlabeled data. However, it often suffers from label inconsistency or limited diversity, which leads to poor performance. In this work, we prove that the core reason for this is lack of a clustering-friendly property in the embedding space. We address this by minimizing the inter- to intra-class similarity ratio to provide clustering-friendly embedding features, and validate our approach t…
A key assumption in supervised learning is that training and test data follow the same probability distribution. However, this fundamental assumption is not always satisfied in practice, e.g., due to changing environments, sample selection bias, privacy concerns, or high labeling costs. Transfer learning (TL) relaxes this assumption and allows us to learn under distribution shift. Classical TL methods typically rely on importance-weighting -- a predictor is trained based on the training losses weighted according to the importance (i.e., the test-over-training density ratio). However, as real-w…
Content and style (C-S) disentanglement intends to decompose the underlying explanatory factors of objects into two independent subspaces. From the unsupervised disentanglement perspective, we rethink content and style and propose a formulation for unsupervised C-S disentanglement based on our assumption that different factors are of different importance and popularity for image reconstruction, which serves as a data bias. The corresponding model inductive bias is introduced by our proposed C-S disentanglement Module (C-S DisMo), which assigns different and independent roles to content and sty…
Is Meta-Learning Out? Rethinking Unsupervised Few-Shot Classification with Limited Entropy
Huazhong University of Science and Technology（华中科技大学）；Nanyang Technological University（南洋理工大学）；University of Washington（华盛顿大学）；University of Copenhagen（哥本哈根大学）
Active learning typically focuses on training a model on few labeled examples alone, while unlabeled ones are only used for acquisition. In this work we depart from this setting by using both labeled and unlabeled data during model training across active learning cycles. We do so by using unsupervised feature learning at the beginning of the active learning pipeline and semi-supervised learning at every active learning cycle, on all available data. The former has not been investigated before in active learning, while the study of latter in the context of deep learning is scarce and recent find…
Since hardware resources are limited, the objective of training deep learning models is typically to maximize accuracy subject to the time and memory constraints of training and inference. We study the impact of model size in this setting, focusing on Transformer models for NLP tasks that are limited by compute: self-supervised pretraining and high-resource machine translation. We first show that even though smaller Transformer models execute faster per iteration, wider and deeper models converge in significantly fewer steps. Moreover, this acceleration in convergence typically outpaces the ad…
Rethinking Tokenizer and Decoder in Masked Graph Modeling for Molecules
National University of Singapore（新加坡国立大学）；University of Science and Technology of China（中国科学技术大学）；Hokkaido University（北海道大学）；Institute of Artificial Intelligence, Institute of Dataspace, Hefei Comprehensive National Science Center（合肥综合性国家科学中心人工智能研究院、数据空间研究院）
