_id stringlengths 4 10 | text stringlengths 0 18.4k | title stringlengths 0 8.56k |
|---|---|---|
d36483539 | Model compression is significant for wide adoption of Recurrent Neural Networks (RNNs) in both user devices possessing limited resources and in business clusters requiring quick responses to large-scale service requests. In this work, we focus on reducing the sizes of basic structures (including input updates, gates, hidden states, cell states and outputs) within Long Short-Term Memory (LSTM) units, so as to learn structurally-sparse LSTMs. Independently reducing the sizes of those basic structures can result in unmatched dimensions among them, and consequently, end up with invalid LSTM units. To overcome this, we propose Intrinsic Sparse Structures (ISS) in LSTMs. By reducing one component of ISS, the sizes of those basic structures are simultaneously reduced by one such that the consistency of dimensions is maintained. By learning ISS within LSTM units, the eventual LSTMs are still regular LSTMs but have much smaller sizes of basic structures. Our method achieves 10.59× speedup in state-of-the-art LSTMs, without losing any perplexity of language modeling of Penn TreeBank dataset. It is also successfully evaluated through a compact model with only 2.69M weights for machine Question Answering of SQuAD dataset. Our source code is public available 1 . | LEARNING INTRINSIC SPARSE STRUCTURES WITHIN LONG SHORT-TERM MEMORY |
d256662499 | Reliability of machine learning evaluation -the consistency of observed evaluation scores across replicated model training runs -is affected by several sources of nondeterminism which can be regarded as measurement noise.Current tendencies to remove noise in order to enforce reproducibility of research results neglect inherent nondeterminism at the implementation level and disregard crucial interaction effects between algorithmic noise factors and data properties.This limits the scope of conclusions that can be drawn from such experiments.Instead of removing noise, we propose to incorporate several sources of variance, including their interaction with data properties, into an analysis of significance and reliability of machine learning evaluation, with the aim to draw inferences beyond particular instances of trained models.We show how to use linear mixed effects models (LMEMs) to analyze performance evaluation scores, and to conduct statistical inference with a generalized likelihood ratio test (GLRT).This allows us to incorporate arbitrary sources of noise like meta-parameter variations into statistical significance testing, and to assess performance differences conditional on data properties.Furthermore, a variance component analysis (VCA) enables the analysis of the contribution of noise sources to overall variance and the computation of a reliability coefficient by the ratio of substantial to total variance. | TOWARDS INFERENTIAL REPRODUCIBILITY OF MACHINE LEARNING RESEARCH |
d263333942 | The capacity of a modern deep learning system to determine if a sample falls within its realm of knowledge is fundamental and important.In this paper, we offer insights and analyses of recent state-of-the-art out-of-distribution (OOD) detection methods -extremely simple activation shaping (ASH).We demonstrate that activation pruning has a detrimental effect on OOD detection, while activation scaling enhances it.Moreover, we propose SCALE, a simple yet effective posthoc network enhancement method for OOD detection, which attains state-ofthe-art OOD detection performance without compromising in-distribution (ID) accuracy.By integrating scaling concepts into the training process to capture a sample's ID characteristics, we propose Intermediate Tensor SHaping (ISH), a lightweight method for training time OOD detection enhancement.We achieve AUROC scores of +1.85% for near-OOD and +0.74% for far-OOD datasets on the OpenOOD v1.5 ImageNet-1K benchmark.Our code and models are available at https://github.com/kai422/SCALE. | SCALING FOR TRAINING TIME AND POST-HOC OUT-OF-DISTRIBUTION DETECTION ENHANCEMENT |
d249712234 | A major challenge in modern machine learning is theoretically understanding the generalization properties of overparameterized models. Many existing tools rely on uniform convergence (UC), a property that, when it holds, guarantees that the test loss will be close to the training loss, uniformly over a class of candidate models. Nagarajan and Kolter (2019b) show that in certain simple linear and neural-network settings, any uniform convergence bound will be vacuous, leaving open the question of how to prove generalization in settings where UC fails. Our main contribution is proving novel generalization bounds in two such settings, one linear, and one non-linear. We study the linear classification setting of Nagarajan and Kolter (2019b), and a quadratic ground truth function learned via a two-layer neural network in the non-linear regime. We prove a new type of margin bound showing that above a certain signal-to-noise threshold, any near-max-margin classifier will achieve almost no test loss in these two settings. Our results show that near-max-margin is important: while any model that achieves at least a (1 − )-fraction of the max-margin generalizes well, a classifier achieving half of the max-margin may fail terribly. Building on the impossibility results of Nagarajan and Kolter (2019b), under slightly stronger assumptions, we show that one-sided UC bounds and classical margin bounds will fail on near-max-margin classifiers. Our analysis provides insight on why memorization can coexist with generalization: we show that in this challenging regime where generalization occurs but UC fails, near-max-margin classifiers simultaneously contain some generalizable components and some overfitting components that memorize the data. The presence of the overfitting components is enough to preclude UC, but the near-extremal margin guarantees that sufficient generalizable components are present.Our work achieves results in two settings. The first is a linear setting previously studied by Nagarajan and Kolter (2019b) where both the ground truth and the trained model are linear. In the second nonlinear setting, studied before by Wei et al.(2019);Frei et al. (2022b), the ground truth is quadratic, and the trained model is a two-layer neural network. In both settings, the data is drawn from a product distribution on features involved in the ground truth labeling function, and "junk" features orthogonal to the signal. We formalize the two settings below.Linear settingData Distribution. Fix some ground truth unit vector direction µ ∈ R d . Let x = z + ξ, where z ∼ Uniform({µ, −µ}) and ξ is uniform on the sphere of radius √ d − 1σ in d − 1 dimensions, orthogonal to the direction µ. Let y = µ T x, such that y = 1 with probability 1/2 and −1 with probability 1/2. We denote this distribution of (Setting for Two-Layer Neural Network Model with Quadratic "XOR" Ground Truth Data Distribution. Fix some orthogonal ground truth unit vector directions µ 1 and µ 2 in R d . Let x = z + ξ, where z ∼ Uniform({µ 1 , −µ 1 , µ 2 , −µ 2 }) and ξ is uniform on the sphere of radius √ d − 2σ in d − 2 dimensions, orthogonal to the directions µ 1 and µ 2 . Let y = (µ T 1 x) 2 − (µ T 2 x) 2 for some orthogonal ground truth directions µ 1 and µ 2 (seeFigure 2(left)). We denote this distribution of (x, y) on R d × {−1, 1} by D µ1,µ2,σ,d . We call this the XOR problem because y = XOR (µ 1 + µ 2 ) T x, (−µ 1 + µ 2 ) T x . For instance, if µ 1 = e 1 and µ 2 = e 2 , then y = x 2 1 − x 2 2 . As can be seen inFigure 2(left), this distribution is not linearly separable, and so one must use nonlinear model to learn in this setting.Model. Fix a ∈ {−1, 1} m so that i a i = 0. The model is a two-layer neural network with m hidden units and activation function φ, parameterized by W ∈ R m×d . W (which will be learned) represents the weights of the first layer and a (which is fixed) is the second layer weights. The model predicts, where w i ∈ R d denotes the i'th column of W . We work with activations φ of the form φ(z) = max(0, z) h for h ∈ [1, 2), and require that m is divisible by 4 1 .We define a problem class of distributions to be a set of data distributions. In this paper, we work with the linear problem class Ω linear σ,d := {D µ,σ,d : µ ∈ R d , µ = 1}, and the quadratic problem class Ω XOR σ,d := {D µ1,µ2,σ,d : µ 1 ⊥ µ 2 ∈ R d , µ 1 = µ 2 = 1}. Here · denotes the 2 norm. We will sometimes abuse notation and say that x ∼ D instead of saying that (x, y) ∼ D.Before proceeding, we make some comments on our models and compare it to related work.Large dimension assumption. In both the linear and non-linear settings, our focus is an overparameterized regime where the dimension d is at least a constant factor times larger than n, the number of training samples. Such an assumption is mild relative to the assumptions made in related work, which require d = ω(n) (see Remark 3.2 for a detailed discussion of work on linear models; for neural networks, the work of Frei et al. 1 The assumption that m is divisible by 4 is for convenience, and can be removed if m is large enough. | Max-Margin Works while Large Margin Fails: Generalization without Uniform Convergence |
d67749672 | Many machine learning algorithms represent input data with vector embeddings or discrete codes. When inputs exhibit compositional structure (e.g. objects built from parts or procedures from subroutines), it is natural to ask whether this compositional structure is reflected in the the inputs' learned representations. While the assessment of compositionality in languages has received significant attention in linguistics and adjacent fields, the machine learning literature lacks general-purpose tools for producing graded measurements of compositional structure in more general (e.g. vector-valued) representation spaces. We describe a procedure for evaluating compositionality by measuring how well the true representation-producing model can be approximated by a model that explicitly composes a collection of inferred representational primitives. We use the procedure to provide formal and empirical characterizations of compositional structure in a variety of settings, exploring the relationship between compositionality and learning dynamics, human judgments, representational similarity, and generalization.We are left with a need for a standard, formal, automatable and quantitative technique for evaluating claims about compositional structure in learned representations. The present work aims at first steps toward meeting that need. We focus on an oracle setting where the compositional structure of model inputs is known, and where the only question is whether this structure is reflected in model outputs. This oracle evaluation paradigm covers most of the existing representation learning problems in which compositionality has been studied.The first contribution of this paper is a simple formal framework for measuring how well a collection of representations (discrete-or continuous-valued) reflects an oracle compositional analysis of model inputs. We propose an evaluation metric called TRE, which provides graded judgments of compositionality for a given set of (input, representation) pairs. The core of our proposal is to treat a set of primitive meaning representations as hidden, and optimize over them to find an explicitly compositional model that approximates the true model as well as possible. For example, if the compositional structure that describes an object is a simple conjunction of attributes, we can search for a collection of "attribute vectors" that sum together to produce the observed object representations; if it is a sparse combination of (attribute, value) pairs we can additionally search for "value vectors" and parameters of a binding operation; and so on for more complex compositions. | MEASURING COMPOSITIONALITY IN REPRESENTATION LEARNING |
d85501306 | In this paper, we propose a residual non-local attention network for high-quality image restoration. Without considering the uneven distribution of information in the corrupted images, previous methods are restricted by local convolutional operation and equal treatment of spatial-and channel-wise features. To address this issue, we design local and non-local attention blocks to extract features that capture the long-range dependencies between pixels and pay more attention to the challenging parts. Specifically, we design trunk branch and (non-)local mask branch in each (non-)local attention block. The trunk branch is used to extract hierarchical features. Local and non-local mask branches aim to adaptively rescale these hierarchical features with mixed attentions. The local mask branch concentrates on more local structures with convolutional operations, while non-local attention considers more about long-range dependencies in the whole feature map. Furthermore, we propose residual local and non-local attention learning to train the very deep network, which further enhance the representation ability of the network. Our proposed method can be generalized for various image restoration applications, such as image denoising, demosaicing, compression artifacts reduction, and super-resolution. Experiments demonstrate that our method obtains comparable or better results compared with recently leading methods quantitatively and visually. Stacked denoising auto-encoder (Vincent et al., 2008) is one of the best well-known CNN based model for image restoration. Dong et al. proposed SRCNN (Dong et al., 2014) for image superresolution and ARCNN (Dong et al., 2015) for image compression artifacts reduction. Both SRCNN and ARCNN achieved superior performance against previous works. By introducing residual learning to ease the training difficulty for deeper network, Zhang et al. proposed DnCNN (Zhang et al., 2017a) for image denoising and compression artifacts reduction. The denoiser prior was lately introduced in IRCNN (Zhang et al., 2017b) for fast image restoration. Mao et al. proposed a very deep fully convolutional encoding-decoding framework with symmetric skip connections for image restoration (Mao et al., 2016). Tai et al. later proposed a very deep end-to-end persistent memory network (MemNet) for image restoration (Tai et al., 2017) and achieved promising results. These CNN based methods have demonstrated the great ability of CNN for image restoration tasks.1 However, there are mainly three issues in the existing CNN based methods above. First, the receptive field size of these networks is relatively small. Most of them extract features in a local way with convolutional operation, which fails to capture the long-range dependencies between pixels in the whole image. A larger receptive field size allows to make better use of training inputs and more context information. This would be very helpful to capture the latent degradation model of LQ images, especially when the images suffer from heavy corruptions. Second, distinctive ability of these networks is also limited. Let's take image denoising as an example. For a noisy image, the noise may appear in both the plain and textural regions. Noise removal would be easier in the plain area than that in the textural one. It is desired to make the denoising model focus on textual area more. However, most previous denoising methods neglect to consider different contents in the noisy input and treat them equally. This would result in over-smoothed outputs and some textural details would also fail to be recovered. Third, all channel-wise features are treated equally in those networks. This naive treatment lacks flexibility in dealing with different types of information (e.g., low-and high-frequency information). For a set of features, some contain more information related to HQ image and the others may contain more information related to corruptions. The interdependencies among channels should be considered for more accurate image restoration.To address the above issues, we propose the very deep residual non-local attention networks (RNAN) for high-quality image restoration. We design residual local and non-local attention blocks as the basic building modules for the very deep network. Each attention block consists of trunk and mask branches. We introduce residual block for trunk branch and extract hierarchical features. For mask branch, we conduct feature downscaling and upscaling with largestride convolution and deconvolution to enlarge receptive field size. Furthermore, we incorporate non-local block in the mask branch to obtain residual non-local mixed attention. We apply RNAN for various restoration tasks, including image denoising, demosaicing, and compression artifacts reduction. Extensive experiments show that our proposed RNAN achieves state-of-the-art results compared with other recent leading methods in all tasks. To the best of our knowledge, this is the first time to consider residual non-local attention for image restoration problems.The main contributions of this work are three-fold:• We propose the very deep residual non-local networks for high-quality image restoration.The powerful networks are based on our proposed residual local and non-local attention blocks, which consist of trunk and mask branches. The network obtains non-local mixed attention with non-local block in the mask branch. Such attention mechanis helps to learn local and non-local information from the hierarchical features.• We propose residual non-local attention learning to train very deep networks by preserving more low-level features, being more suitable for image restoration. Using non-local lowlevel and high-level attention from the very deep network, we can pursue better network representational ability and finally obtain high-quality image restoration results.• We demonstrate with extensive experiments that our RNAN is powerful for various image restoration tasks. RNAN achieves superior results over leading methods for image denoising, demosaicing, compression artifacts reduction, and super-resolution. In addition, RNAN achieves superior performance with moderate model size and performs very fast.RELATED WORKNon-local prior. As a classical filtering algorithm, non-local means (Buades et al., 2005) is computed as weighted mean of all pixels of an image. Such operation allows distant pixels to contribute to the response of a position at a time. It was lately introduced in BM3D (Dabov et al., 2007b) for image denoising. Recently, Wang et al. (2018a) proposed non-local neural network by incorporating non-local operations in deep neural network for video classification. We can see that those methods mainly introduce non-local information in the trunk pipeline. Liu et al.(2018)proposed non-local recurrent network for image restoration. However, in this paper, we mainly focus on learning nonlocal attention to better guide feature extraction in trunk branch.Attention mechanisms. Generally, attention can be viewed as a guidance to bias the allocation of available processing resources towards the most informative components of an input (Hu et al., representations. In Proc. 7th Int. Conf. Curves Surf., 2010. . Ffdnet: Toward a fast and flexible solution for cnn based image denoising. arXiv preprint arXiv:1710.04026, 2017c. . Learning a single convolutional super-resolution network for multiple degradations. In CVPR, 2018a. | RESIDUAL NON-LOCAL ATTENTION NETWORKS FOR IMAGE RESTORATION |
d43933271 | Coding theory is a central discipline underpinning wireline and wireless modems that are the workhorses of the information age. Progress in coding theory is largely driven by individual human ingenuity with sporadic breakthroughs over the past century. In this paper we study whether it is possible to automate the discovery of decoding algorithms via deep learning. We study a family of sequential codes parametrized by recurrent neural network (RNN) architectures. We show that creatively designed and trained RNN architectures can decode well known sequential codes such as the convolutional and turbo codes with close to optimal performance on the additive white Gaussian noise (AWGN) channel, which itself is achieved by breakthrough algorithms of our times (Viterbi and BCJR decoders, representing dynamic programing and forward-backward algorithms). We show strong generalizations, i.e., we train at a specific signal to noise ratio and block length but test at a wide range of these quantities, as well as robustness and adaptivity to deviations from the AWGN setting. | Communication Algorithms via Deep Learning |
d29153681 | Adapting deep networks to new concepts from a few examples is challenging, due to the high computational requirements of standard fine-tuning procedures.Most work on few-shot learning has thus focused on simple learning techniques for adaptation, such as nearest neighbours or gradient descent. Nonetheless, the machine learning literature contains a wealth of methods that learn non-deep models very efficiently. In this paper, we propose to use these fast convergent methods as the main adaptation mechanism for few-shot learning. The main idea is to teach a deep network to use standard machine learning tools, such as ridge regression, as part of its own internal model, enabling it to quickly adapt to novel data. This requires back-propagating errors through the solver steps. While normally the cost of the matrix operations involved in such a process would be significant, by using the Woodbury identity we can make the small number of examples work to our advantage. We propose both closed-form and iterative solvers, based on ridge regression and logistic regression components. Our methods constitute a simple and novel approach to the problem of few-shot learning and achieve performance competitive with or superior to the state of the art on three benchmarks. | META-LEARNING WITH DIFFERENTIABLE CLOSED-FORM SOLVERS |
d44065372 | The success of popular algorithms for deep reinforcement learning, such as policygradients and Q-learning, relies heavily on the availability of an informative reward signal at each timestep of the sequential decision-making process. When rewards are only sparsely available during an episode, or a rewarding feedback is provided only after episode termination, these algorithms perform sub-optimally due to the difficultly in credit assignment. Alternatively, trajectory-based policy optimization methods, such as cross-entropy method and evolution strategies, do not require per-timestep rewards, but have been found to suffer from high sample complexity by completing forgoing the temporal nature of the problem. Improving the efficiency of RL algorithms in real-world problems with sparse or episodic rewards is therefore a pressing need. In this work, we introduce a self-imitation learning algorithm that exploits and explores well in the sparse and episodic reward settings. We view each policy as a state-action visitation distribution and formulate policy optimization as a divergence minimization problem. We show that with Jensen-Shannon divergence, this divergence minimization problem can be reduced into a policy-gradient algorithm with shaped rewards learned from experience replays. Experimental results indicate that our algorithm works comparable to existing algorithms in environments with dense rewards, and significantly better in environments with sparse and episodic rewards. We then discuss limitations of self-imitation learning, and propose to solve them by using Stein variational policy gradient descent with the Jensen-Shannon kernel to learn multiple diverse policies. We demonstrate its effectiveness on a challenging variant of continuous-control MuJoCo locomotion tasks. | LEARNING SELF-IMITATING DIVERSE POLICIES |
d232307426 | This paper introduces Relative Predictive Coding (RPC), a new contrastive representation learning objective that maintains a good balance among training stability, minibatch size sensitivity, and downstream task performance. The key to the success of RPC is two-fold. First, RPC introduces the relative parameters to regularize the objective for boundedness and low variance. Second, RPC contains no logarithm and exponential score functions, which are the main cause of training instability in prior contrastive objectives. We empirically verify the effectiveness of RPC on benchmark vision and speech self-supervised learning tasks. Lastly, we relate RPC with mutual information (MI) estimation, showing RPC can be used to estimate MI with low variance 1 . | SELF-SUPERVISED REPRESENTATION LEARNING WITH RELATIVE PREDICTIVE CODING |
d244908617 | Denoising diffusion probabilistic models have recently received much research attention since they outperform alternative approaches, such as GANs, and currently provide state-of-the-art generative performance. The superior performance of diffusion models has made them an appealing tool in several applications, including inpainting, super-resolution, and semantic editing. In this paper, we demonstrate that diffusion models can also serve as an instrument for semantic segmentation, especially in the setup when labeled data is scarce. In particular, for several pretrained diffusion models, we investigate the intermediate activations from the networks that perform the Markov step of the reverse diffusion process. We show that these activations effectively capture the semantic information from an input image and appear to be excellent pixel-level representations for the segmentation problem. Based on these observations, we describe a simple segmentation method, which can work even if only a few training images are provided. Our approach significantly outperforms the existing alternatives on several datasets for the same amount of human supervision. The source code of the project is publicly available.Published as a conference paper at ICLR 2022 2. We design a simple semantic segmentation approach that exploits these representations and outperforms the alternatives in the few-shot operating point.3. We compare the DDPM-based representations with their GAN-based counterparts on the same datasets and demonstrate the advantages of the former in the context of semantic segmentation.RELATED WORKIn this section, we briefly describe the existing lines of research relevant to our work.Diffusion models(Sohl-Dickstein et al., 2015;Ho et al., 2020)are a class of generative models that approximate the distribution of real images by the endpoint of the Markov chain which originates from a simple parametric distribution, typically a standard Gaussian. Each Markov step is modeled by a deep neural network that effectively learns to invert the diffusion process with a known Gaussian kernel. Ho et al. highlighted the equivalence of diffusion models and score matching(Song & Ermon, 2019;, showing them to be two different perspectives on the gradual conversion of a simple known distribution into a target distribution via the iterative denoising process. Very recent works(Nichol, 2021;Dhariwal & Nichol, 2021)have developed more powerful model architectures as well as different advanced objectives, which led to the "victory" of DDPM over GANs in terms of generative quality and diversity. DDPM have been widely used in several applications, including image colorization ), super-resolution (Saharia et al., 2021Li et al., 2021b), inpainting (Song et al., 2021, and semantic editing(Meng et al., 2021). In our work, we demonstrate that one can also successfully use them for semantic segmentation. | LABEL-EFFICIENT SEMANTIC SEGMENTATION WITH DIFFUSION MODELS |
d253265328 | By conditioning on natural language instructions, large language models (LLMs) have displayed impressive capabilities as general-purpose computers. However, task performance depends significantly on the quality of the prompt used to steer the model, and most effective prompts have been handcrafted by humans. Inspired by classical program synthesis and the human approach to prompt engineering, we propose Automatic Prompt Engineer 1 (APE) for automatic instruction generation and selection. In our method, we treat the instruction as the "program," optimized by searching over a pool of instruction candidates proposed by an LLM in order to maximize a chosen score function. To evaluate the quality of the selected instruction, we evaluate the zero-shot performance of another LLM following the selected instruction. Extensive experiments show that our automatically generated instructions outperform the prior LLM baseline by a large margin and achieve better or comparable performance to the instructions generated by human annotators on 24/24 Instruction Induction tasks and 17/21 curated BIG-Bench tasks. We conduct extensive qualitative and quantitative analyses to explore the performance of APE. We show that APE-engineered prompts are able to improve few-shot learning performance (by simply prepending them to standard in-context learning prompts), find better zero-shot chain-ofthought prompts, as well as steer models toward truthfulness and/or informativeness. 2 , et al. Program synthesis with large language models. arXiv preprint arXiv:2108.07732, 2021. , et al. Evaluating large language models trained on code. arXiv preprint arXiv:2107.03374, 2021. : Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805, 2018. Discarded tasks: abstraction and reasoning corpus, bbq lite, bias from probabilities, boolean expressions, com2sense, context definition alignment, convinceme, coqa conversational question answering, cycled letters, diverse social bias, dynamic counting, factuality of summary, forecasting subquestions, gender sensitivity chinese, gender sensitivity english, high low game, long context integration, multistep arithmetic, muslim violence bias, program synthesis, protein interacting sites, python programming challenge, question answer creation, roots optimization and games, self awareness, self evaluation courtroom, self evaluation tutoring, simple arithmetic, spelling bee, squad shifts, subject verb agreement, sudoku, taboo, talkdown, text navigation game, training on test set, truthful qa, twenty questions, unqover, web of lies, word problems on sets and graphs, yes no black white.Criteria: Tasks without sub-task. Discarded tasks: abstract narrative understanding, arithmetic, authorship verification, bbq lite json, cause and effect, chess state tracking, cifar10 classification, color, conceptual combinations, conlang translation, cs algorithms, elementary math qa, fact checker, gem, goal step wikihow, hhh alignment, indic cause and effect, intersect geometry, kanji ascii, key value maps, language games, linguistic mappings, list functions, logical deduction, metaphor understanding, minute mysteries qa, modified arithmetic, mult data wrangling, multiemo, natural instructions, periodic elements, physics, real or fake text, simp turing concept, simple arithmetic json subtasks, simple ethical questions, strange stories, symbol interpretation, tracking shuffled objects, undo permutation, unit conversion, unit interpretation, unnatural in context learning.Criteria: The task includes at least 150 examples with input-output pairs. Discarded tasks: analytic entailment, auto debugging, code line description, codenames, common morpheme, crash blossom, crass ai, cryobiology spanish, dark humor detection, emoji movie, emojis emotion prediction, empirical judgments, english proverbs, english russian proverbs, entailed polarity, entailed polarity hindi, evaluating information essentiality, figure of speech detection, general knowledge, gre reading comprehension, human organs senses, identify math theorems, identify odd metaphor, implicit relations, international phonetic alphabet nli, irony identification, known unknowns, logical args, logical sequence, mathematical induction, misconceptions russian, nonsense words grammar, novel concepts, odd one out, penguins in a table, persian idioms, phrase relatedness, physical intuition, physics questions, repeat copy logic, rephrase, riddle sense, scientific press release, sentence ambiguity, similarities abstraction, simple arithmetic json, simple arithmetic json multiple choice, simple arithmetic multiple targets json, simple text editing, sufficient information, suicide risk, swedish to german proverbs, what is the tao.Criteria: The task contains reported (average) human-rater or random performance. Discarded tasks: contextual parametric knowledge conflicts, hinglish toxicity, medical questions russian, parsinlu qa, swahili english proverbs, tellmewhy, which wiki edit.Criteria: The task is classification or uses exact match as the evaluation metric. Discarded tasks: auto categorization, few shot nlg, hindi question answering, international phonetic alphabet transliterate, polish sequence labeling, qa wikidata, semantic parsing in context sparc, semantic parsing spider, social support, topical chat.Published as a conference paper at ICLR 2023 | LARGE LANGUAGE MODELS ARE HUMAN-LEVEL PROMPT ENGINEERS |
d238531695 | Transformers have made great progress in dealing with computer vision tasks.However, existing vision transformers do not yet possess the ability of building the interactions among features of different scales, which is perceptually important to visual inputs.The reasons are two-fold: (1) Input embeddings of each layer are equal-scale, so no cross-scale feature can be extracted; (2) to lower the computational cost, some vision transformers merge adjacent embeddings inside the selfattention module, thus sacrificing small-scale (fine-grained) features of the embeddings and also disabling the cross-scale interactions.To this end, we propose Cross-scale Embedding Layer (CEL) and Long Short Distance Attention (LSDA).On the one hand, CEL blends each embedding with multiple patches of different scales, providing the self-attention module itself with cross-scale features.On the other hand, LSDA splits the self-attention module into a short-distance one and a long-distance counterpart, which not only reduces the computational burden but also keeps both small-scale and large-scale features in the embeddings.Through the above two designs, we achieve cross-scale attention.Besides, we put forward a dynamic position bias for vision transformers to make the popular relative position bias apply to variable-sized images.Hinging on the cross-scale attention module, we construct a versatile vision architecture, dubbed CrossFormer, which accommodates variable-sized inputs.Extensive experiments show that CrossFormer outperforms the other vision transformers on image classification, object detection, instance segmentation, and semantic segmentation tasks. 1 | CROSSFORMER: A VERSATILE VISION TRANSFORMER HINGING ON CROSS-SCALE ATTENTION |
d263830433 | Zero-shot learning in prompted vision-language models, the practice of crafting prompts to build classifiers without an explicit training process, has achieved impressive performance in many settings.This success presents a seemingly surprising observation: these methods suffer relatively little from overfitting, i.e., when a prompt is manually engineered to achieve low error on a given training set (thus rendering the method no longer actually zero-shot), the approach still performs well on held-out test data.In this paper, we show that we can explain such performance well via recourse to classical PAC-Bayes bounds.Specifically, we show that the discrete nature of prompts, combined with a PAC-Bayes prior given by a language model, results in generalization bounds that are remarkably tight by the standards of the literature: for instance, the generalization bound of an ImageNet classifier is often within a few percentage points of the true test error.We demonstrate empirically that this holds for existing handcrafted prompts and prompts generated through simple greedy search.Furthermore, the resulting bound is well-suited for model selection: the models with the best bound typically also have the best test performance.This work thus provides a possible justification for the widespread practice of "prompt engineering," even if it seems that such methods could potentially overfit the training data. | Understanding prompt engineering may not require rethinking generalization |
d254408645 | Variational autoencoders (VAEs) are powerful tools for learning latent representations of data used in a wide range of applications. In practice, VAEs usually require multiple training rounds to choose the amount of information the latent variable should retain. This trade-off between the reconstruction error (distortion) and the KL divergence (rate) is typically parameterized by a hyperparameter β. In this paper, we introduce Multi-Rate VAE (MR-VAE), a computationally efficient framework for learning optimal parameters corresponding to various β in a single training run. The key idea is to explicitly formulate a response function that maps β to the optimal parameters using hypernetworks. MR-VAEs construct a compact response hypernetwork where the pre-activations are conditionally gated based on β. We justify the proposed architecture by analyzing linear VAEs and showing that it can represent response functions exactly for linear VAEs. With the learned hypernetwork, MR-VAEs can construct the rate-distortion curve without additional training and can be deployed with significantly less hyperparameter tuning. Empirically, our approach is competitive and often exceeds the performance of multiple β-VAEs training with minimal computation and memory overheads. | Multi-Rate VAE: Train Once, Get the Full Rate-Distortion Curve |
d3464416 | The problem of detecting whether a test sample is from in-distribution (i.e., training distribution by a classifier) or out-of-distribution sufficiently different from it arises in many real-world machine learning applications. However, the state-of-art deep neural networks are known to be highly overconfident in their predictions, i.e., do not distinguish in-and out-of-distributions. Recently, to handle this issue, several threshold-based detectors have been proposed given pre-trained neural classifiers. However, the performance of prior works highly depends on how to train the classifiers since they only focus on improving inference procedures. In this paper, we develop a novel training method for classifiers so that such inference algorithms can work better. In particular, we suggest two additional terms added to the original loss (e.g., cross entropy). The first one forces samples from out-of-distribution less confident by the classifier and the second one is for (implicitly) generating most effective training samples for the first one. In essence, our method jointly trains both classification and generative neural networks for out-of-distribution. We demonstrate its effectiveness using deep convolutional neural networks on various popular image datasets. | TRAINING CONFIDENCE-CALIBRATED CLASSIFIERS FOR DETECTING OUT-OF-DISTRIBUTION SAMPLES |
d252693131 | Despite the tremendous success, existing machine learning models still fall short of human-like systematic generalization-learning compositional rules from limited data and applying them to unseen combinations in various domains. We propose Neural-Symbolic Recursive Machine (NSR) to tackle this deficiency. The core representation of NSR is a Grounded Symbol System (GSS) with combinatorial syntax and semantics, which entirely emerges from training data. Akin to the neuroscience studies suggesting separate brain systems for perceptual, syntactic, and semantic processing, NSR implements analogous separate modules of neural perception, syntactic parsing, and semantic reasoning, which are jointly learned by a deduction-abduction algorithm. We prove that NSR is expressive enough to model various sequence-to-sequence tasks. Superior systematic generalization is achieved via the inductive biases of equivariance and recursiveness embedded in NSR. In experiments, NSR achieves state-of-the-art performance in three benchmarks from different domains: SCAN for semantic parsing, PCFG for string manipulation, and HINT for arithmetic reasoning. Specifically, NSR achieves 100% generalization accuracy on SCAN and PCFG and outperforms state-of-the-art models on HINT by about 23%. Our NSR demonstrates stronger generalization than pure neural networks due to its symbolic representation and inductive biases. NSR also demonstrates better transferability than existing neural-symbolic approaches due to less domain-specific knowledge required. | NEURAL-SYMBOLIC RECURSIVE MACHINE FOR SYSTEMATIC GENERALIZATION |
d235359206 | We consider sketching algorithms which first compress data by multiplication with a random sketch matrix, and then apply the sketch to quickly solve an optimization problem, e.g., low-rank approximation and regression. In the learning-based sketching paradigm proposed by [IVY19], the sketch matrix is found by choosing a random sparse matrix, e.g., CountSketch, and then the values of its non-zero entries are updated by running gradient descent on a training data set. Despite the growing body of work on this paradigm, a noticeable omission is that the locations of the non-zero entries of previous algorithms were fixed, and only their values were learned. In this work, we propose the first learning-based algorithms that also optimize the locations of the non-zero entries. Our first proposed algorithm is based on a greedy algorithm. However, one drawback of the greedy algorithm is its slower training time. We fix this issue and propose approaches for learning a sketching matrix for both low-rank approximation and Hessian approximation for second order optimization. The latter is helpful for a range of constrained optimization problems, such as LASSO and matrix estimation with a nuclear norm constraint. Both approaches achieve good accuracy with a fast running time. Moreover, our experiments suggest that our algorithm can still reduce the error significantly even if we only have a very limited number of training matrices.algorithms for CUDA, a widely used GPU library. They investigated the (classical) sketched singular value decomposition (SVD), but found that the solutions were not accurate enough across a spectrum of inputs [CB19]. This is precisely the issue addressed by the learned sketch paradigm where we optimize for "good" average performance across a range of inputs.While promising results have been shown using previous learned sketching techniques, notable gaps remain. In particular, all previous methods work by initializing the sketching matrix with a random sparse matrix, e.g., each column of the sketching matrix has a single non-zero value chosen at a uniformly random position. Then, the values of the non-zero entries are updated by running gradient descent on a training data set, or via other methods. However, the locations of the non-zero entries are held fixed throughout the entire training process.Clearly this is sub-optimal. Indeed, suppose the input matrix A is an n × d matrix with first d rows equal to the d × d identity matrix, and remaining rows equal to 0. A random sketching matrix S with a single non-zero per column is known to require m = Ω(d 2 ) rows in order for S · A to preserve the rank of A [NN14]; this follows by a birthday paradox argument. On the other hand, it is clear that if S is a d × n matrix with first d rows equal to the identity matrix, then ∥S · Ax∥ 2 = ∥Ax∥ 2 for all vectors x, and so S preserves not only the rank of A but all important spectral properties. A random matrix would be very unlikely to choose the non-zero entries in the first d columns of S so perfectly, whereas an algorithm trained to optimize the locations of the non-zero entries would notice and correct for this. This is precisely the gap in our understanding that we seek to fill.Learned CountSketch Paradigm of [IVY19]. Throughout the paper, we assume our data A ∈ R n×d is sampled from an unknown distribution D. Specifically, we have a training setThe generic form of our optimization problems is min X f (A, X), where A ∈ R n×d is the input matrix. For a given optimization problem and a set S of sketching matrices, define ALG(S, A) to be the output of the classical sketching algorithm resulting from using S; this uses the sketching matrices in S to map the given input A and construct an approximate solution X. We remark that the number of sketches used by an algorithm can vary and in its simplest case, S is a single sketch, but in more complicated sketching approaches we may need to apply sketching more than once-hence S may also denote a set of more than one sketching matrix.The learned sketch framework has two parts: (1) offline sketch learning and (2) "online" sketching (i.e., applying the learned sketch and some sketching algorithm to possibly unseen data). In offline sketch learning, the goal is to construct a CountSketch matrix (abbreviated as CS matrix) with the minimum expected error for the problem of interest. Formally, that is,where X * denotes the optimal solution. Moreover, the minimum is taken over all possible constructions of CS. We remark that when ALG needs more than one CS to be learned (e.g., in the sketching algorithm we consider for LRA), we optimize each CS independently using a surrogate loss function.In the second part of the learned sketch paradigm, we take the sketch from part one and use it within a sketching algorithm. This learned sketch and sketching algorithm can be applied, again and again, to different inputs. Finally, we augment the sketching algorithm to provide worst-case guarantees when used with learned sketches. The goal is to have good performance on A ∈ D while the worst-case performance on A ̸ ∈ D remains comparable to the guarantees of classical sketches. We remark that the learned matrix S is trained offline only once using the training data. Hence, no additional computational cost is incurred when solving the optimization problem on the test data. | Learning the Positions in CountSketch |
d209485520 | Optimistic initialisation is an effective strategy for efficient exploration in reinforcement learning (RL). In the tabular case, all provably efficient model-free algorithms rely on it. However, model-free deep RL algorithms do not use optimistic initialisation despite taking inspiration from these provably efficient tabular algorithms. In particular, in scenarios with only positive rewards, Q-values are initialised at their lowest possible values due to commonly used network initialisation schemes, a pessimistic initialisation. Merely initialising the network to output optimistic Q-values is not enough, since we cannot ensure that they remain optimistic for novel state-action pairs, which is crucial for exploration. We propose a simple count-based augmentation to pessimistically initialised Q-values that separates the source of optimism from the neural network. We show that this scheme is provably efficient in the tabular setting and extend it to the deep RL setting. Our algorithm, Optimistic Pessimistically Initialised Q-Learning (OPIQ), augments the Q-value estimates of a DQN-based agent with count-derived bonuses to ensure optimism during both action selection and bootstrapping. We show that OPIQ outperforms non-optimistic DQN variants that utilise a pseudocount-based intrinsic motivation in hard exploration tasks, and that it predicts optimistic estimates for novel state-action pairs.Published as a conference paper at ICLR 2020BACKGROUNDWe consider a Markov Decision Process (MDP) defined as a tuple (S, A, P, R), where S is the state space, A is the discrete action space, P (·|s, a) is the state-transition distribution, R(·|s, a) is the distribution over rewards and γ ∈ [0, 1) is the discount factor. The goal of the agent is then to maximise the expected discounted sum of rewards: E[ ∞ t=0 γ t r t |r t ∼ R(·|s t , a t )], in the discounted episodic setting. A policy π(·|s) is a mapping from states to actions such that it is a valid probability distribution. Q π (s, a) := E[ ∞ t=0 γ t r t |a t ∼ π(·|s t )] and Q * := max π Q π . Deep Q-Network (DQN) (Mnih et al., 2015) uses a nonlinear function approximator (a deep neural network) to estimate the action-value function, Q(s, a; θ) ≈ Q * (s, a), where θ are the parameters of the network. Exploration based on intrinsic rewards (e.g.,Bellemare et al., 2016), which uses a DQN agent, additionally augments the observed rewards r t with a bonus β/ N (s t , a t ) based on Published as a conference paper at ICLR 2020 | OPTIMISTIC EXPLORATION EVEN WITH A PESSIMISTIC INITIALISATION |
d67855732 | We propose a single neural probabilistic model based on variational autoencoder that can be conditioned on an arbitrary subset of observed features and then sample the remaining features in "one shot". The features may be both real-valued and categorical. Training of the model is performed by stochastic variational Bayes. The experimental evaluation on synthetic data, as well as feature imputation and image inpainting problems, shows the effectiveness of the proposed approach and diversity of the generated samples.Published as a conference paper at ICLR 2019The experimental evaluation shows that the proposed model successfully samples from the conditional distributions. The distribution over samples is close to the true conditional distribution. This property is very important when the true distribution has several modes. The model is shown to be effective in feature imputation problem which helps to increase the quality of subsequent discriminative models on different problems from UCI datasets collection (Lichman, 2013). We demonstrate that model can generate diverse and realistic image inpaintings on MNIST (LeCun et al., 1998), Omniglot (Lake et al., 2015 and CelebA (Liu et al., 2015) datasets, and works even better than the current state of the art inpainting techniques in terms of peak signal to noise ratio (PSNR).The paper is organized as follows. In section 2 we review the related works. In section 3 we briefly describe variational autoencoders and conditional variational autoencoders. In section 4 we define the problem, describe the VAEAC model and its training procedure. In section 5 we evaluate VAEAC. Section 6 concludes the paper. Appendix contains additional explanations, theoretical analysis, and experiments for VAEAC. 2 RELATED WORK Universal Marginalizer (Douglas et al., 2017) is a model based on a feed-forward neural network which approximates marginals of unobserved features conditioned on observable values. A related idea of an autoregressive model of joint probability was previously proposed in Germain et al. (2015). The detailed description of the model, experimental evaluation and comparison with VAEAC are contained in section 5.3. Yoon et al. (2018) propose a GANs-based model called GAIN which solves the same problem as VAEAC.In contrast to VAEAC, GAIN doesnt use unobserved data during training, which makes it easier to apply to the missing features imputation problem. Nevertheless, it turns into a disadvantage when the fully-observed training data is available but the missingness rate at the testing stage is high. For example, in inpainting setting GAIN cannot learn the conditional distribution over MNIST digits given one horizontal line of the image while VAEAC can (see appendix D.4). The comparison of VAEAC and GAIN on the missing feature imputation problem is given in section 5.1 and appendix D.2. Rezende et al. (2014) [Appendix F], Sohl-Dickstein et al. (2015), Goyal et al. (2017), and Bordes et al. (2017) propose to fill missing data with noise and run Markov chain with a learned transition operator. The stationary distribution of such chains approximates the true conditional distribution of the unobserved features. Nevertheless, such methods are computationally expensive at the test time and require fully-observed training data.Image inpainting is a classic computer vision problem. Most of the earlier methods rely on local and texture information or hand-crafted problem-specific features(Bertalmio et al., 2000). In past years multiple neural network based approaches have been proposed. Pathak et al. (2016), Yeh et al. (2016) and Yang et al. (2017) use different kinds and combinations of adversarial, reconstruction, texture and other losses. Li et al. (2017) focuses on face inpainting and uses two adversarial losses and one semantic parsing loss to train the generative model. In Yeh et al. (2017)GANs are first trained on the whole training dataset. The inpainting is an optimization procedure that finds the latent variables that explain the observed features best. Then, the obtained latents are passed through the generative model to restore the unobserved portion of the image. We can say that VAEAC is a similar model which uses prior network to find a proper latents instead of solving the optimization problem.All described methods aim to produce a single realistic inpainting, while VAEAC is capable of sampling diverse inpaintings. Additionally, Yeh et al. (2016), Yang et al. (2017) and Yeh et al. (2017 have high testtime computational complexity of inpainting, because they require an optimization problem to be solved. On the other hand, VAEAC is a "single-shot" method with a low computational cost. | VARIATIONAL AUTOENCODER WITH ARBITRARY CONDITIONING |
d256390109 | Despite the popularity and efficacy of knowledge distillation, there is limited understanding of why it helps. In order to study the generalization behavior of a distilled student, we propose a new theoretical framework that leverages supervision complexity: a measure of alignment between teacher-provided supervision and the student's neural tangent kernel. The framework highlights a delicate interplay among the teacher's accuracy, the student's margin with respect to the teacher predictions, and the complexity of the teacher predictions. Specifically, it provides a rigorous justification for the utility of various techniques that are prevalent in the context of distillation, such as early stopping and temperature scaling. Our analysis further suggests the use of online distillation, where a student receives increasingly more complex supervision from teachers in different stages of their training. We demonstrate efficacy of online distillation and validate the theoretical findings on a range of image classification benchmarks and model architectures. . On kernel-target alignment. Advances in neural information processing systems, 14, 2001.Tri Dao, Govinda M Kamath, Vasilis Syrgkanis, and Lester Mackey. Knowledge distillation as semiparametric inference. . A linearized framework and a new benchmark for model selection for fine-tuning. arXiv preprint arXiv:2102.00084, 2021. | Supervision Complexity and its Role in Knowledge Distillation |
d221818900 | In this work, we propose DiffWave, a versatile Diffusion probabilistic model for conditional and unconditional Waveform generation. The model is nonautoregressive, and converts the white noise signal into structured waveform through a Markov chain with a constant number of steps at synthesis. It is efficiently trained by optimizing a variant of variational bound on the data likelihood. DiffWave produces high-fidelity audios in Different Waveform generation tasks, including neural vocoding conditioned on mel spectrogram, class-conditional generation, and unconditional generation. We demonstrate that DiffWave matches a strong WaveNet vocoder in terms of speech quality (MOS: 4.44 versus 4.43), while synthesizing orders of magnitude faster. In particular, it significantly outperforms autoregressive and GAN-based waveform models in the challenging unconditional generation task in terms of audio quality and sample diversity from various automatic and human evaluations. | DIFFWAVE: A VERSATILE DIFFUSION MODEL FOR AUDIO SYNTHESIS |
d246430471 | In nature, the process of cellular growth and differentiation has lead to an amazing diversity of organisms -algae, starfish, giant sequoia, tardigrades, and orcas are all created by the same generative process. Inspired by the incredible diversity of this biological generative process, we propose a generative model, the Variational Neural Cellular Automata (VNCA), which is loosely inspired by the biological processes of cellular growth and differentiation. Unlike previous related works, the VNCA is a proper probabilistic generative model, and we evaluate it according to best practices. We find that the VNCA learns to reconstruct samples well and that despite its relatively few parameters and simple local-only communication, the VNCA can learn to generate a large variety of output from information encoded in a common vector format. While there is a significant gap to the current state-of-the-art in terms of generative modeling performance, we show that the VNCA can learn a purely self-organizing generative process of data. Additionally, we show that the VNCA can learn a distribution of stable attractors that can recover from significant damage. | VARIATIONAL NEURAL CELLULAR AUTOMATA |
d238583193 | We propose a new class of random feature methods for linearizing softmax and Gaussian kernels called hybrid random features (HRFs) that automatically adapt the quality of kernel estimation to provide most accurate approximation in the defined regions of interest. Special instantiations of HRFs lead to well-known methods such as trigonometric (Rahimi & Recht, 2007) or (recently introduced in the context of linear-attention Transformers) positive random features (Choromanski et al., 2021b). By generalizing Bochner's Theorem for softmax/Gaussian kernels and leveraging random features for compositional kernels, the HRFmechanism provides strong theoretical guarantees -unbiased approximation and strictly smaller worst-case relative errors than its counterparts. We conduct exhaustive empirical evaluation of HRF ranging from pointwise kernel estimation experiments, through tests on data admitting clustering structure to benchmarking implicit-attention Transformers (also for downstream Robotics applications), demonstrating its quality in a wide spectrum of machine learning problems. * Equal Contribution, Correspondence to kchoro@google.com. † Authorship in alphabetical order ‡ Independent researcher. Work done during postdoc at Yale University arXiv:2110.04367v3 [cs.LG] 30 Jan 2022 2 HYBRID RANDOM FEATURES 2.1 PRELIMINARIES Whenever we do not say explicitly otherwise, all presented lemmas and theorems are new. We start with the following basic definitions and results.Definition 2.1 (Kernel with a Random Feature Map Representation). We say that a kernel function K : R d × R d → R admits a random feature (RF) map representation if it can be written as | HYBRID RANDOM FEATURES |
d247996510 | The field of Natural Language Processing (NLP) has experienced a dramatic leap in capabilities with the recent introduction of huge Language Models (LMs). Despite this success, natural language problems that involve several compounded steps are still practically unlearnable, even by the largest LMs. This complies with experimental failures for end-to-end learning of composite problems that were demonstrated in a variety of domains. An effective mitigation is to introduce intermediate supervision for solving sub-tasks of the compounded problem.Recently, several works have demonstrated high gains by taking a straightforward approach for incorporating intermediate supervision in compounded natural language problems: the sequence-to-sequence LM is fed with an augmented input, in which the decomposed tasks' labels are simply concatenated to the original input (seefigure 1). In this paper, we prove a positive learning result that motivates these recent efforts. We show that when concatenating intermediate supervision to the input and training a sequence-to-sequence model on this modified input, unlearnable composite problems can become learnable. We show that this is true for any family of tasks which on the one hand, are unlearnable, and on the other hand, can be decomposed into a polynomial number of simple sub-tasks, each of which depends only on O(1) previous sub-task results. Beyond motivating contemporary empirical efforts for incorporating intermediate supervision in sequence-to-sequence language models, our positive theoretical result is the first of its kind in the landscape of results on the benefits of intermediate supervision for neural-network learning: Until now, all theoretical results on the subject are negative, i.e., show cases where learning is impossible without intermediate supervision, while our result is positive, showing that learning is facilitated in the presence of intermediate supervision. | SUB-TASK DECOMPOSITION ENABLES LEARNING IN SEQUENCE TO SEQUENCE TASKS |
d226965123 | When designing controllers for safety-critical systems, practitioners often face a challenging tradeoff between robustness and performance. While robust control methods provide rigorous guarantees on system stability under certain worst-case disturbances, they often result in simple controllers that perform poorly in the average (non-worst) case. In contrast, nonlinear control methods trained using deep learning have achieved state-of-the-art performance on many control tasks, but often lack robustness guarantees. In this paper, we propose a technique that combines the strengths of these two approaches: a generic nonlinear control policy class, parameterized by neural networks, that nonetheless enforces the same provable robustness criteria as robust control. Specifically, we show that by integrating custom convex-optimization-based projection layers into a nonlinear policy, we can construct a provably robust neural network policy class that outperforms robust control methods in the average (non-adversarial) setting. We demonstrate the power of this approach on several domains, improving in performance over existing robust control methods and in stability over (non-robust) RL methods. | Enforcing robust control guarantees within neural network policies |
d257279836 | Preference-based reinforcement learning (RL) provides a framework to train agents using human preferences between two behaviors.However, preferencebased RL has been challenging to scale since it requires a large amount of human feedback to learn a reward function aligned with human intent.In this paper, we present Preference Transformer, a neural architecture that models human preferences using transformers.Unlike prior approaches assuming human judgment is based on the Markovian rewards which contribute to the decision equally, we introduce a new preference model based on the weighted sum of non-Markovian rewards.We then design the proposed preference model using a transformer architecture that stacks causal and bidirectional self-attention layers.We demonstrate that Preference Transformer can solve a variety of control tasks using real human preferences, while prior approaches fail to work.We also show that Preference Transformer can induce a well-specified reward and attend to critical events in the trajectory by automatically capturing the temporal dependencies in human decision-making.Code is available on the project website: https://sites.google.com/view/preference-transformer. | PREFERENCE TRANSFORMER: MODELING HUMAN PREFERENCES USING TRANSFORMERS FOR RL |
d203838029 | The capability of making interpretable and self-explanatory decisions is essential for developing responsible machine learning systems. In this work, we study the learning to explain problem in the scope of inductive logic programming (ILP). We propose Neural Logic Inductive Learning (NLIL), an efficient differentiable ILP framework that learns first-order logic rules that can explain the patterns in the data. In experiments, compared with the state-of-the-art methods, we find NLIL can search for rules that are x10 times longer while remaining x3 times faster. We also show that NLIL can scale to large image datasets, i.e. Visual Genome, with 1M entities. | LEARN TO EXPLAIN EFFICIENTLY VIA NEURAL LOGIC INDUCTIVE LEARNING |
d236087352 | A desirable property of autonomous agents is the ability to both solve long-horizon problems and generalize to unseen tasks. Recent advances in data-driven skill learning have shown that extracting behavioral priors from offline data can enable agents to solve challenging long-horizon tasks with reinforcement learning. However, generalization to tasks unseen during behavioral prior training remains an outstanding challenge. To this end, we present Few-shot Imitation with Skill Transition Models (FIST), an algorithm that extracts skills from offline data and utilizes them to generalize to unseen tasks given a few downstream demonstrations. FIST learns an inverse skill dynamics model, a distance function, and utilizes a semi-parametric approach for imitation. We show that FIST is capable of generalizing to new tasks and substantially outperforms prior baselines in navigation experiments requiring traversing unseen parts of a large maze and 7-DoF robotic arm experiments requiring manipulating previously unseen objects in a kitchen. arXiv:2107.08981v2 [cs.LG] | HIERARCHICAL FEW-SHOT IMITATION WITH SKILL TRANSITION MODELS |
d257757426 | With the advent of large datasets, offline reinforcement learning (RL) is a promising framework for learning good decision-making policies without the need to interact with the real environment. However, offline RL requires the dataset to be reward-annotated, which presents practical challenges when reward engineering is difficult or when obtaining reward annotations is labor-intensive. In this paper, we introduce Optimal Transport Reward labeling (OTR), an algorithm that assigns rewards to offline trajectories, with a few high-quality demonstrations. OTR's key idea is to use optimal transport to compute an optimal alignment between an unlabeled trajectory in the dataset and an expert demonstration to obtain a similarity measure that can be interpreted as a reward, which can then be used by an offline RL algorithm to learn the policy. OTR is easy to implement and computationally efficient. On D4RL benchmarks, we show that OTR with a single demonstration can consistently match the performance of offline RL with ground-truth rewards 1 . | OPTIMAL TRANSPORT FOR OFFLINE IMITATION LEARNING |
d239769065 | Goal-conditioned reinforcement learning (RL) can solve tasks in a wide range of domains, including navigation and manipulation, but learning to reach distant goals remains a central challenge to the field. Learning to reach such goals is particularly hard without any offline data, expert demonstrations, and reward shaping. In this paper, we propose an algorithm to solve the distant goal-reaching task by using search at training time to automatically generate a curriculum of intermediate states. Our algorithm, Classifier-Planning (C-Planning), frames the learning of the goal-conditioned policies as expectation maximization: the E-step corresponds to planning an optimal sequence of waypoints using graph search, while the M-step aims to learn a goal-conditioned policy to reach those waypoints. Unlike prior methods that combine goal-conditioned RL with graph search, ours performs search only during training and not testing, significantly decreasing the compute costs of deploying the learned policy. Empirically, we demonstrate that our method is more sample efficient that prior methods. Moreover, it is able to solve very long horizons manipulation and navigation tasks, tasks that prior goalconditioned methods and methods based on graph search fail to solve. | C-PLANNING: AN AUTOMATIC CURRICULUM FOR LEARNING GOAL-REACHING TASKS |
d248810913 | A large number of studies that analyze deep neural network models and their ability to encode various linguistic and non-linguistic concepts provide an interpretation of the inner mechanics of these models. The scope of the analyses is limited to pre-defined concepts that reinforce the traditional linguistic knowledge and do not reflect on how novel concepts are learned by the model. We address this limitation by discovering and analyzing latent concepts learned in neural network models in an unsupervised fashion and provide interpretations from the model's perspective. In this work, we study: i) what latent concepts exist in the pre-trained BERT model, ii) how the discovered latent concepts align or diverge from classical linguistic hierarchy and iii) how the latent concepts evolve across layers. Our findings show: i) a model learns novel concepts (e.g. animal categories and demographic groups), which do not strictly adhere to any pre-defined categorization (e.g. POS, semantic tags), ii) several latent concepts are based on multiple properties which may include semantics, syntax, and morphology, iii) the lower layers in the model dominate in learning shallow lexical concepts while the higher layers learn semantic relations and iv) the discovered latent concepts highlight potential biases learned in the model. We also release 1 a novel BERT ConceptNet dataset (BCN) consisting of 174 concept labels and 1M annotated instances. * Equal contribution 1 Code and dataset: https://neurox.qcri.org/projects/bert-concept-net.html | DISCOVERING LATENT CONCEPTS LEARNED IN BERT |
d52881014 | We present a meta-learning approach for adaptive text-to-speech (TTS) with few data. During training, we learn a multi-speaker model using a shared conditional WaveNet core and independent learned embeddings for each speaker. The aim of training is not to produce a neural network with fixed weights, which is then deployed as a TTS system. Instead, the aim is to produce a network that requires few data at deployment time to rapidly adapt to new speakers. We introduce and benchmark three strategies: (i) learning the speaker embedding while keeping the WaveNet core fixed, (ii) fine-tuning the entire architecture with stochastic gradient descent, and (iii) predicting the speaker embedding with a trained neural network encoder. The experiments show that these approaches are successful at adapting the multi-speaker neural network to new speakers, obtaining state-of-the-art results in both sample naturalness and voice similarity with merely a few minutes of audio data from new speakers. | SAMPLE EFFICIENT ADAPTIVE TEXT-TO-SPEECH |
d1091965 | Convolutional Neural Networks (CNNs) are computationally intensive, which limits their application on mobile devices. Their energy is dominated by the number of multiplies needed to perform the convolutions. Winograd's minimal filtering algorithm(Lavin, 2015)and network pruning(Han et al., 2015)can reduce the operation count, but these two methods cannot be directly combined -applying the Winograd transform fills in the sparsity in both the weights and the activations. We propose two modifications to Winograd-based CNNs to enable these methods to exploit sparsity. First, we move the ReLU operation into the Winograd domain to increase the sparsity of the transformed activations. Second, we prune the weights in the Winograd domain to exploit static weight sparsity. For models on CIFAR-10, CIFAR-100 and ImageNet datasets, our method reduces the number of multiplications by 10.4×, 6.8× and 10.8× respectively with loss of accuracy less than 0.1%, outperforming previous baselines by 2.0×-3.0×. We also show that moving ReLU to the Winograd domain allows more aggressive pruning. | EFFICIENT SPARSE-WINOGRAD CONVOLUTIONAL NEURAL NETWORKS |
d218901036 | The vulnerability of deep networks to adversarial attacks is a central problem for deep learning from the perspective of both cognition and security. The current most successful defense method is to train a classifier using adversarial images created during learning. Another defense approach involves transformation or purification of the original input to remove adversarial signals before the image is classified. We focus on defending naturally-trained classifiers using Markov Chain Monte Carlo (MCMC) sampling with an Energy-Based Model (EBM) for adversarial purification. In contrast to adversarial training, our approach is intended to secure pre-existing and highly vulnerable classifiers. The memoryless behavior of long-run MCMC sampling will eventually remove adversarial signals, while metastable behavior preserves consistent appearance of MCMC samples after many steps to allow accurate long-run prediction. Balancing these factors can lead to effective purification and robust classification. We evaluate adversarial defense with an EBM using the strongest known attacks against purification. Our contributions are 1) an improved method for training EBM's with realistic long-run MCMC samples, 2) an Expectation-Over-Transformation (EOT) defense that resolves theoretical ambiguities for stochastic defenses and from which the EOT attack naturally follows, and 3) state-of-the-art adversarial defense for naturally-trained classifiers and competitive defense compared to adversariallytrained classifiers on Cifar-10, SVHN, and Cifar-100. Code and pre-trained models are available at https://github.com/point0bar1/ebm-defense. | Stochastic Security: Adversarial Defense Using Long-Run Dynamics of Energy-Based Models |
d237593074 | We present Pix2Seq, a simple and generic framework for object detection. Unlike existing approaches that explicitly integrate prior knowledge about the task, we cast object detection as a language modeling task conditioned on the observed pixel inputs. Object descriptions (e.g., bounding boxes and class labels) are expressed as sequences of discrete tokens, and we train a neural network to perceive the image and generate the desired sequence. Our approach is based mainly on the intuition that if a neural network knows about where and what the objects are, we just need to teach it how to read them out. Beyond the use of task-specific data augmentations, our approach makes minimal assumptions about the task, yet it achieves competitive results on the challenging COCO dataset, compared to highly specialized and well optimized detection algorithms. 1 | PIX2SEQ: A LANGUAGE MODELING FRAMEWORK FOR OBJECT DETECTION |
d257079026 | Given a matrix M ∈ R m×n , the low rank matrix completion problem asks us to find a rank-k approximation of M as U V ⊤ for U ∈ R m×k and V ∈ R n×k by only observing a few entries specified by a set of entries Ω ⊆ [m] × [n]. In particular, we examine an approach that is widely used in practice -the alternating minimization framework. Jain, Netrapalli and Sanghavi [JNS13] showed that if M has incoherent rows and columns, then alternating minimization provably recovers the matrix M by observing a nearly linear in n number of entries. While the sample complexity has been subsequently improved [GLZ17], alternating minimization steps are required to be computed exactly. This hinders the development of more efficient algorithms and fails to depict the practical implementation of alternating minimization, where the updates are usually performed approximately in favor of efficiency.In this paper, we take a major step towards a more efficient and error-robust alternating minimization framework. To this end, we develop an analytical framework for alternating minimization that can tolerate moderate amount of errors caused by approximate updates. Moreover, our algorithm runs in time O(|Ω|k), which is nearly linear in the time to verify the solution while preserving the sample complexity. This improves upon all prior known alternating minimization approaches which require O(|Ω|k 2 ) | Low Rank Matrix Completion via Robust Alternating Minimization in Nearly Linear Time |
d238583808 | We study model-based offline Reinforcement Learning with general function approximation. We present an algorithm named Constrained Pessimistic Policy Optimization (CPPO) which leverages a general function class and uses a constraint to encode pessimism. Under the assumption that the ground truth model belongs to our function class, CPPO can learn with the offline data only providing partial coverage, i.e., it can learn a policy that competes against any policy that is covered by the offline data, in polynomial sample complexity with respect to the statistical complexity of the function class. We then demonstrate that this algorithmic framework can be applied to many specialized Markov Decision Processes and the additional structural assumptions can further refine the concept of partial coverage. One notable example is low-rank MDP with representation learning where the partial coverage is defined using the concept of relative condition number measured by the underlying unknown ground truth feature representation. Finally, we introduce and study the Bayesian setting in offline RL.The key benefit of Bayesian offline RL is that algorithmically, we do not need to explicitly construct pessimism or reward penalty which could be hard beyond models with linear structures. We present a posterior sampling based incremental policy optimization algorithm (PS-PO) which proceeds by iteratively sampling a model from the posterior distribution and performing one step incremental policy optimization inside the sampled model. Theoretically, in expectation with respect to the prior distribution, PS-PO can learn a near optimal policy under partial coverage with polynomial sample complexity. This work is a long version of the conference paper in https://openreview.net/pdf?id=tyrJsbKAe6. | Pessimistic Model-based Offline Reinforcement Learning under Partial Coverage |
d235358707 | Self-supervised learning has recently shown great potential in vision tasks through contrastive learning, which aims to discriminate each image, or instance, in the dataset. However, such instance-level learning ignores the semantic relationship among instances and sometimes undesirably repels the anchor from the semantically similar samples, termed as "false negatives". In this work, we show that the unfavorable effect from false negatives is more significant for the large-scale datasets with more semantic concepts. To address the issue, we propose a novel self-supervised contrastive learning framework that incrementally detects and explicitly removes the false negative samples. Specifically, following the training process, our method dynamically detects increasing high-quality false negatives considering that the encoder gradually improves and the embedding space becomes more semantically structural. Next, we discuss two strategies to explicitly remove the detected false negatives during contrastive learning. Extensive experiments show that our framework outperforms other self-supervised contrastive learning methods on multiple benchmarks in a limited resource setup. The source code is available at https://github.com/tsaishien-chen/IFND. | INCREMENTAL FALSE NEGATIVE DETECTION FOR CONTRASTIVE LEARNING |
d248377648 | In this work, we evaluate the effectiveness of representation learning approaches for decision making in visually complex environments. Representation learning is essential for effective reinforcement learning (RL) from high-dimensional inputs. Unsupervised representation learning approaches based on reconstruction, prediction or contrastive learning have shown substantial learning efficiency gains. Yet, they have mostly been evaluated in clean laboratory or simulated settings. In contrast, real environments are visually complex and contain substantial amounts of clutter and distractors. Unsupervised representations will learn to model such distractors, potentially impairing the agent's learning efficiency. In contrast, an alternative class of approaches, which we call task-induced representation learning, leverages task information such as rewards or demonstrations from prior tasks to focus on task-relevant parts of the scene and ignore distractors. We investigate the effectiveness of unsupervised and task-induced representation learning approaches on four visually complex environments, from Distracting DMControl to the CARLA driving simulator. For both, RL and imitation learning, we find that representation learning generally improves sample efficiency on unseen tasks even in visually complex scenes and that task-induced representations can double learning efficiency compared to unsupervised alternatives. | TASK-INDUCED REPRESENTATION LEARNING |
d210932183 | One-shot neural architecture search (NAS) has played a crucial role in making NAS methods computationally feasible in practice. Nevertheless, there is still a lack of understanding on how these weight-sharing algorithms exactly work due to the many factors controlling the dynamics of the process. In order to allow a scientific study of these components, we introduce a general framework for one-shot NAS that can be instantiated to many recently-introduced variants and introduce a general benchmarking framework that draws on the recent large-scale tabular benchmark NAS-Bench-101 for cheap anytime evaluations of one-shot NAS methods. To showcase the framework, we compare several state-of-the-art one-shot NAS methods, examine how sensitive they are to their hyperparameters and how they can be improved by tuning their hyperparameters, and compare their performance to that of blackbox optimizers for NAS-Bench-101. | NAS-BENCH-1SHOT1: BENCHMARKING AND DISSECTING ONE-SHOT NEURAL ARCHITECTURE SEARCH |
d49667227 | How can we build a learner that can capture the essence of what makes a hard problem more complex than a simple one, break the hard problem along characteristic lines into smaller problems it knows how to solve, and sequentially solve the smaller problems until the larger one is solved? To work towards this goal, we focus on learning to generalize in a particular family of problems that exhibit compositional and recursive structure: their solutions can be found by composing in sequence a set of reusable partial solutions. Our key idea is to recast the problem of generalization as a problem of learning algorithmic procedures: we can formulate a solution to this family as a sequential decision-making process over transformations between representations. Our formulation enables the learner to learn the structure and parameters of its own computation graph with sparse supervision, make analogies between problems by transforming one problem representation to another, and exploit modularity and reuse to scale to problems of varying complexity. Experiments on solving a variety of multilingual arithmetic problems demonstrate that our method discovers the hierarchical decomposition of a problem into its subproblems, generalizes out of distribution to unseen problem classes, and extrapolates to harder versions of the same problem, yielding a 10-fold reduction in sample complexity compared to a monolithic recurrent neural network. | Automatically Composing Representation Transformations as a Means for Generalization |
d219792972 | Unsupervised meta-learning approaches rely on synthetic meta-tasks that are created using techniques such as random selection, clustering and/or augmentation. Unfortunately, clustering and augmentation are domain-dependent, and thus they require either manual tweaking or expensive learning. In this work, we describe an approach that generates meta-tasks using generative models. A critical component is a novel approach of sampling from the latent space that generates objects grouped into synthetic classes forming the training and validation data of a metatask. We find that the proposed approach, LAtent Space Interpolation Unsupervised Meta-learning (LASIUM), outperforms or is competitive with current unsupervised learning baselines on few-shot classification tasks on the most widely used benchmark datasets. In addition, the approach promises to be applicable without manual tweaking over a wider range of domains than previous approaches.IntroductionMeta-learning algorithms for neural networks (1; 2; 3) prepare networks to quickly adapt to unseen tasks. This is done in a meta-training phase that typically involves a large number of supervised learning tasks. Very recently, several approaches had been proposed that perform the meta-training by generating synthetic training tasks from an unsupervised dataset. This requires us to generate samples with specific pairwise information: in-class pairs of samples that are with high likelihood in the same class, and out-of-class pairs that are with high likelihood not in the same class. For instance, UMTRA (4) and AAL (5) achieve this through random selection from a domain with many classes for out-of-class pairs and by augmentation for in-class pairs. CACTUs (6) creates synthetic labels through unsupervised clustering of the domain. Unfortunately, these algorithms depend on domain specific expertise for the appropriate clustering and augmentation techniques.In this paper, we rely on recent advances in the field of generative models, such as the variants of generative adversarial networks (GANs) and variational autoencoders (VAEs), to generate the in-class and out-of-class pairs of meta-training data. The fundamental idea of our approach is that in-class pairs are close while out-of-class pairs are far away in the latent space representation of the generative model. Thus, we can generate in-class pairs by interpolating between two out-of-class samples * These authors contributed equally.Preprint. Under review. | Unsupervised Meta-Learning through Latent-Space Interpolation in Generative Models |
d235352772 | We developed Distilled Graph Attention Policy Network (DGAPN), a reinforcement learning model to generate novel graph-structured chemical representations that optimize user-defined objectives by efficiently navigating a physically constrained domain. The framework is examined on the task of generating molecules that are designed to bind, noncovalently, to functional sites of SARS-CoV-2 proteins. We present a spatial Graph Attention (sGAT) mechanism that leverages self-attention over both node and edge attributes as well as encoding the spatial structure -this capability is of considerable interest in synthetic biology and drug discovery. An attentional policy network is introduced to learn the decision rules for a dynamic, fragment-based chemical environment, and state-of-the-art policy gradient techniques are employed to train the network with stability. Exploration is driven by the stochasticity of the action space design and the innovation reward bonuses learned and proposed by random network distillation. In experiments, our framework achieved outstanding results compared to state-of-the-art algorithms, while reducing the complexity of paths to chemical synthesis.Published as a conference paper at ICLR 2022 dynamic and chemically valid fragment-based action space. We also propose a hybrid Graph Neural Network (GNN) that comprehensively encodes graph objects' attributes and spatial structures in addition to adjacency structures. The following paragraphs discuss how we addressed limitations of prior work and its relevance to antiviral drug discovery. For more descriptions of key prior methodologies that we used as benchmarks in this paper, see Section 4. | SPATIAL GRAPH ATTENTION AND CURIOSITY-DRIVEN POLICY FOR ANTIVIRAL DRUG DISCOVERY |
d253098484 | Many applications of quantum computing in the near term rely on variational quantum circuits (VQCs). They have been showcased as a promising model for reaching a quantum advantage in machine learning with current noisy intermediate scale quantum computers (NISQ). It is often believed that the power of VQCs relies on their exponentially large feature space, and extensive works have explored the expressiveness and trainability of VQCs in that regard. In our work, we propose a classical sampling method that may closely approximate a VQC with Hamiltonian encoding, given only the description of its architecture. It uses the seminal proposal of Random Fourier Features (RFF) and the fact that VQCs can be seen as large Fourier series. We provide general theoretical bounds for classically approximating models built from exponentially large quantum feature space by sampling a few frequencies to build an equivalent low dimensional kernel, and we show experimentally that this approximation is efficient for several encoding strategies. Precisely, we show that the number of required samples grows favorably with the size of the quantum spectrum. This tool therefore questions the hope for quantum advantage from VQCs in many cases, but conversely helps to narrow the conditions for their potential success. We expect VQCs with various and complex encoding Hamiltonians, or with large input dimension, to become more robust to classical approximations. | Classically Approximating Variational Quantum Machine Learning with Random Fourier Features |
d256105646 | Federated Averaging (FedAvg) remains the most popular algorithm for Federated Learning (FL) optimization due to its simple implementation, stateless nature, and privacy guarantees combined with secure aggregation.Recent work has sought to generalize the vanilla averaging in FedAvg to a generalized gradient descent step by treating client updates as pseudo-gradients and using a server step size.While the use of a server step size has been shown to provide performance improvement theoretically, the practical benefit of the server step size has not been seen in most existing works.In this work, we present FedExP, a method to adaptively determine the server step size in FL based on dynamically varying pseudo-gradients throughout the FL process.We begin by considering the overparameterized convex regime, where we reveal an interesting similarity between FedAvg and the Projection Onto Convex Sets (POCS) algorithm.We then show how FedExP can be motivated as a novel extension to the extrapolation mechanism that is used to speed up POCS.Our theoretical analysis later also discusses the implications of FedExP in underparameterized and non-convex settings.Experimental results show that FedExP consistently converges faster than FedAvg and competing baselines on a range of realistic FL datasets. | FEDEXP: SPEEDING UP FEDERATED AVERAGING VIA EXTRAPOLATION |
d258988005 | We introduce a theoretical framework for sampling from unnormalized densities based on a smoothing scheme that uses an isotropic Gaussian kernel with a single fixed noise scale. We prove one can decompose sampling from a density (minimal assumptions made on the density) into a sequence of sampling from log-concave conditional densities via accumulation of noisy measurements with equal noise levels. Our construction is unique in that it keeps track of a history of samples, making it non-Markovian as a whole, but it is lightweight algorithmically as the history only shows up in the form of a running empirical mean of samples. Our sampling algorithm generalizes walk-jump sampling (Saremi & Hyvärinen, 2019). The "walk" phase becomes a (non-Markovian) chain of (log-concave) Markov chains. The "jump" from the accumulated measurements is obtained by empirical Bayes. We study our sampling algorithm quantitatively using the 2-Wasserstein metric and compare it with various Langevin MCMC algorithms. We also report a remarkable capacity of our algorithm to "tunnel" between modes of a distribution.1 | CHAIN OF LOG-CONCAVE MARKOV CHAINS |
d211171786 | The tremendous success of deep neural networks has motivated the need to better understand the fundamental properties of these networks, but many of the theoretical results proposed have only been for shallow networks. In this paper, we study an important primitive for understanding the meaningful input space of a deep network: span recovery. For k < n, let A ∈ R k×n be the innermost weight matrix of an arbitrary feed forward neural network M : R n → R, so M (x) can be written as M (x) = σ(Ax), for some network σ : R k → R. The goal is then to recover the row span of A given only oracle access to the value of M (x). We show that if M is a multilayered network with ReLU activation functions, then partial recovery is possible: namely, we can provably recover k/2 linearly independent vectors in the row span of A using poly(n) non-adaptive queries to M (x). Furthermore, if M has differentiable activation functions, we demonstrate that full span recovery is possible even when the output is first passed through a sign or 0/1 thresholding function; in this case our algorithm is adaptive. Empirically, we confirm that full span recovery is not always possible, but only for unrealistically thin layers. For reasonably wide networks, we obtain full span recovery on both random networks and networks trained on MNIST data. Furthermore, we demonstrate the utility of span recovery as an attack by inducing neural networks to misclassify data obfuscated by controlled random noise as sensical inputs. | Span Recovery for Deep Neural Networks with Applications to Input Obfuscation |
d221856753 | Gradient descent can be surprisingly good at optimizing deep neural networks without overfitting and without explicit regularization. We find that the discrete steps of gradient descent implicitly regularize models by penalizing gradient descent trajectories that have large loss gradients. We call this Implicit Gradient Regularization (IGR) and we use backward error analysis to calculate the size of this regularization. We confirm empirically that implicit gradient regularization biases gradient descent toward flat minima, where test errors are small and solutions are robust to noisy parameter perturbations. Furthermore, we demonstrate that the implicit gradient regularization term can be used as an explicit regularizer, allowing us to control this gradient regularization directly. More broadly, our work indicates that backward error analysis is a useful theoretical approach to the perennial question of how learning rate, model size, and parameter regularization interact to determine the properties of overparameterized models optimized with gradient descent. * equal contribution . Gradient descent optimizes overparameterized deep relu networks. Machine Learning, 109, 2019. | IMPLICIT GRADIENT REGULARIZATION |
d44005913 | The conventional solution to the recommendation problem greedily ranks individual document candidates by prediction scores. However, this method fails to optimize the slate as a whole, and hence, often struggles to capture biases caused by the page layout and document interdepedencies. The slate recommendation problem aims to directly find the optimally ordered subset of documents (i.e. slates) that best serve users' interests. Solving this problem is hard due to the combinatorial explosion in all combinations of document candidates and their display positions on the page. Therefore we propose a paradigm shift from the traditional viewpoint of solving a ranking problem to a direct slate generation framework. In this paper, we introduce List Conditional Variational Auto-Encoders (List-CVAE), which learns the joint distribution of documents on the slate conditioned on user responses, and directly generates full slates. Experiments on simulated and real-world data show that List-CVAE outperforms popular comparable ranking methods consistently on various scales of documents corpora. | BEYOND GREEDY RANKING: SLATE OPTIMIZATION VIA LIST-CVAE |
d235359192 | The noise in stochastic gradient descent (SGD), caused by minibatch sampling, is poorly understood despite its practical importance in deep learning. This work presents the first systematic study of the SGD noise and fluctuations close to a local minimum. We first analyze the SGD noise in linear regression in detail and then derive a general formula for approximating SGD noise in different types of minima. For application, our results (1) provide insight into the stability of training a neural network, (2) suggest that a large learning rate can help generalization by introducing an implicit regularization, (3) explain why the linear learning ratebatchsize scaling law fails at a large learning rate or at a small batchsize and (4) can provide an understanding of how discrete-time nature of SGD affects the recently discovered power-law phenomenon of SGD.This definition is quite general. Most commonly studied and used loss functions are additive, e.g., the mean-square error (MSE) and cross-entropy loss. For an additive loss, the minibatch SGD with momentum algorithm can be defined. Definition 2. The minibatch SGD with momentum algorithm by sampling with replacement computes the update to the parameter w with the following set of equations:⎧ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎩ĝ t = 1 S ∑ i∈Bt ∇ (x i , y i , w t−1 ); m t = µm t−1 +ĝ t ; w t = w t−1 − λm t . | STRENGTH OF MINIBATCH NOISE IN SGD |
d227054483 | Reinforcement learning provides a general framework for flexible decision making and control, but requires extensive data collection for each new task that an agent needs to learn. In other machine learning fields, such as natural language processing or computer vision, pre-training on large, previously collected datasets to bootstrap learning for new tasks has emerged as a powerful paradigm to reduce data requirements when learning a new task. In this paper, we ask the following question: how can we enable similarly useful pre-training for RL agents? We propose a method for pre-training behavioral priors that can capture complex input-output relationships observed in successful trials from a wide range of previously seen tasks, and we show how this learned prior can be used for rapidly learning new tasks without impeding the RL agent's ability to try out novel behaviors. We demonstrate the effectiveness of our approach in challenging robotic manipulation domains involving image observations and sparse reward functions, where our method outperforms prior works by a substantial margin. Additional materials can be found on our project website: https://sites.google.com/view/parrot-rl. | PARROT: DATA-DRIVEN BEHAVIORAL PRIORS FOR REINFORCEMENT LEARNING |
d238531427 | Multimodal variational autoencoders (VAEs) have shown promise as efficient generative models for weakly-supervised data. Yet, despite their advantage of weak supervision, they exhibit a gap in generative quality compared to unimodal VAEs, which are completely unsupervised. In an attempt to explain this gap, we uncover a fundamental limitation that applies to a large family of mixture-based multimodal VAEs. We prove that the sub-sampling of modalities enforces an undesirable upper bound on the multimodal ELBO and thereby limits the generative quality of the respective models. Empirically, we showcase the generative quality gap on both synthetic and real data and present the tradeoffs between different variants of multimodal VAEs. We find that none of the existing approaches fulfills all desired criteria of an effective multimodal generative model when applied on more complex datasets than those used in previous benchmarks. In summary, we identify, formalize, and validate fundamental limitations of VAE-based approaches for modeling weakly-supervised data and discuss implications for real-world applications. | ON THE LIMITATIONS OF MULTIMODAL VAES |
d235390683 | Graph self-supervised learning has gained increasing attention due to its capacity to learn expressive node representations. Many pretext tasks, or loss functions have been designed from distinct perspectives. However, we observe that different pretext tasks affect downstream tasks differently across datasets, which suggests that searching over pretext tasks is crucial for graph self-supervised learning. Different from existing works focusing on designing single pretext tasks, this work aims to investigate how to automatically leverage multiple pretext tasks effectively. Nevertheless, evaluating representations derived from multiple pretext tasks without direct access to ground truth labels makes this problem challenging. To address this obstacle, we make use of a key principle of many real-world graphs, i.e., homophily, or the principle that "like attracts like," as the guidance to effectively search various self-supervised pretext tasks. We provide theoretical understanding and empirical evidence to justify the flexibility of homophily in this search task. Then we propose the AUTOSSL framework to automatically search over combinations of various self-supervised tasks. By evaluating the framework on 8 real-world datasets, our experimental results show that AUTOSSL can significantly boost the performance on downstream tasks including node clustering and node classification compared with training under individual tasks. | AUTOMATED SELF-SUPERVISED LEARNING FOR GRAPHS |
d244773230 | Complex sequential tasks in continuous-control settings often require agents to successfully traverse a set of "narrow passages" in their state space. Solving such tasks with a sparse reward in a sample-efficient manner poses a challenge to modern reinforcement learning (RL) due to the associated long-horizon nature of the problem and the lack of sufficient positive signal during learning. Various tools have been applied to address this challenge. When available, large sets of demonstrations can guide agent exploration. Hindsight relabelling on the other hand does not require additional sources of information. However, existing strategies explore based on task-agnostic goal distributions, which can render the solution of long-horizon tasks impractical. In this work, we extend hindsight relabelling mechanisms to guide exploration along task-specific distributions implied by a small set of successful demonstrations. We evaluate the approach on four complex, single and dual arm, robotics manipulation tasks against strong suitable baselines. The method requires far fewer demonstrations to solve all tasks and achieves a significantly higher overall performance as task complexity increases. Finally, we investigate the robustness of the proposed solution with respect to the quality of input representations and the number of demonstrations. * Work done during an internship at DeepMind; † Denotes equal contribution. distribution is often considerably more complex and non-stationary in comparison to the target-task distribution. When the final task is very complex and the main task reward is not perceived, this can lead to learning a capable, goal-reaching policy for states close to the agent's initial states while unable to complete the actual task.Our contribution is based on the insight that demonstration states can be viewed as samples of the target task distribution, and can therefore be used to constrain the self-supervision process to task-relevant goals. We introduce a framework for task-constrained goal-conditioned RL that flexibly combines demonstrations with hindsight relabelling. Unlike HER, which learns a general goalconditioned agent, we train a goal-conditioned agent specialized at achieving goals which directly lead to the task solution. The approach further allows to smoothly vary the task relevance of the relabelling process. Unlike conventional goal-conditioned RL, we enable agents to solve tasks through utilising abstract goal formulations, such as inserting a 3.5mm jack, common in the context of complex sequential tasks. We achieve this through using a continuously improving target goal distribution for the online goal selection stage. We refer to our method as a Hindsight Goal Selection for Demo-Driven RL or HinDRL for short. We demonstrate that the proposed solution can solve tasks where both demonstration-driven RL and its self-supervised version with HER struggle. Specifically, we show that HinDRL can dramatically reduce the number of required demonstration by an order of magnitude on most considered tasks. | WISH YOU WERE HERE: HINDSIGHT GOAL SELECTION FOR LONG-HORIZON DEXTEROUS MANIPULATION |
d263828820 | The visual examination of tissue biopsy sections is fundamental for cancer diagnosis, with pathologists analyzing sections at multiple magnifications to discern tumor cells and their subtypes.However, existing attention-based multiple instance learning (MIL) models, used for analyzing Whole Slide Images (WSIs) in cancer diagnostics, often overlook the contextual information of tumor and neighboring tiles, leading to misclassifications.To address this, we propose the Context-Aware Multiple Instance Learning (CAMIL) architecture.CAMIL incorporates neighbor-constrained attention to consider dependencies among tiles within a WSI and integrates contextual constraints as prior knowledge into the MIL model.We evaluated CAMIL on subtyping non-small cell lung cancer (TCGA-NSCLC) and detecting lymph node (CAMELYON16) metastasis, achieving test AUCs of 0.959% and 0.975%, respectively, outperforming other state-of-the-art methods.Additionally, CAMIL enhances model interpretability by identifying regions of high diagnostic value. | CAMIL: Context-Aware Multiple Instance Learning for Cancer Detection and Subtyping in Whole Slide Images |
d249538336 | Recent techniques for approximating Nash equilibria in very large games leverage neural networks to learn approximately optimal policies (strategies). One promising line of research uses neural networks to approximate counterfactual regret minimization (CFR) or its modern variants. DREAM, the only current CFR-based neural method that is model free and therefore scalable to very large games, trains a neural network on an estimated regret target that can have extremely high variance due to an importance sampling term inherited from Monte Carlo CFR (MCCFR). In this paper we propose an unbiased model-free method that does not require any importance sampling. Our method, ESCHER, is principled and is guaranteed to converge to an approximate Nash equilibrium with high probability. We show that the variance of the estimated regret of ESCHER is orders of magnitude lower than DREAM and other baselines. We then show that ESCHER outperforms the prior state of the art-DREAM and neural fictitious self play (NFSP)-on a number of games and the difference becomes dramatic as game size increases. In the very large game of dark chess, ESCHER is able to beat DREAM and NFSP in a head-to-head competition over 90% of the time. | ESCHER: ESCHEWING IMPORTANCE SAMPLING IN GAMES BY COMPUTING A HISTORY VALUE FUNCTION TO ESTIMATE REGRET |
d52893258 | Compression is a key step to deploy large neural networks on resource-constrained platforms. As a popular compression technique, quantization constrains the number of distinct weight values and thus reducing the number of bits required to represent and store each weight. In this paper, we study the representation power of quantized neural networks. First, we prove the universal approximability of quantized ReLU networks on a wide class of functions. Then we provide upper bounds on the number of weights and the memory size for a given approximation error bound and the bit-width of weights for function-independent and functiondependent structures. Our results reveal that, to attain an approximation error bound of , the number of weights needed by a quantized network is no more than O log 5 (1/ ) times that of an unquantized network. This overhead is of much lower order than the lower bound of the number of weights needed for the error bound, supporting the empirical success of various quantization techniques. To the best of our knowledge, this is the first in-depth study on the complexity bounds of quantized neural networks.1 | ON THE UNIVERSAL APPROXIMABILITY AND COMPLEXITY BOUNDS OF QUANTIZED RELU NEURAL NETWORKS |
d54445324 | In complex transfer learning scenarios new tasks might not be tightly linked to previous tasks. Approaches that transfer information contained only in the final parameters of a source model will therefore struggle. Instead, transfer learning at a higher level of abstraction is needed. We propose Leap, a framework that achieves this by transferring knowledge across learning processes. We associate each task with a manifold on which the training process travels from initialization to final parameters and construct a meta-learning objective that minimizes the expected length of this path. Our framework leverages only information obtained during training and can be computed on the fly at negligible cost. We demonstrate that our framework outperforms competing methods, both in meta-learning and transfer learning, on a set of computer vision tasks. Finally, we demonstrate that Leap can transfer knowledge across learning processes in demanding reinforcement learning environments (Atari) that involve millions of gradient steps. * Work done while at Amazon. | TRANSFERRING KNOWLEDGE ACROSS LEARNING PROCESSES |
d59599587 | Deep learning has achieved remarkable successes in solving challenging reinforcement learning (RL) problems when dense reward function is provided. However, in sparse reward environment it still often suffers from the need to carefully shape reward function to guide policy optimization. This limits the applicability of RL in the real world since both reinforcement learning and domain-specific knowledge are required. It is therefore of great practical importance to develop algorithms which can learn from a binary signal indicating successful task completion or other unshaped, sparse reward signals. We propose a novel method called competitive experience replay, which efficiently supplements a sparse reward by placing learning in the context of an exploration competition between a pair of agents. Our method complements the recently proposed hindsight experience replay (HER) by inducing an automatic exploratory curriculum. We evaluate our approach on the tasks of reaching various goal locations in an ant maze and manipulating objects with a robotic arm. Each task provides only binary rewards indicating whether or not the goal is achieved. Our method asymmetrically augments these sparse rewards for a pair of agents each learning the same task, creating a competitive game designed to drive exploration. Extensive experiments demonstrate that this method leads to faster converge and improved task performance. | COMPETITIVE EXPERIENCE REPLAY |
d3290366 | We present graph partition neural networks (GPNN), an extension of graph neural networks (GNNs) able to handle extremely large graphs. GPNNs alternate between locally propagating information between nodes in small subgraphs and globally propagating information between the subgraphs. To efficiently partition graphs, we experiment with several partitioning algorithms and also propose a novel variant for fast processing of large scale graphs. We extensively test our model on a variety of semi-supervised node classification tasks. Experimental results indicate that GPNNs are either superior or comparable to state-of-the-art methods on a wide variety of datasets for graph-based semi-supervised classification. We also show that GPNNs can achieve similar performance as standard GNNs with fewer propagation steps. | Graph Partition Neural Networks for Semi-Supervised Classification |
d254926905 | Oversmoothing is a central challenge of building more powerful Graph Neural Networks (GNNs). While previous works have only demonstrated that oversmoothing is inevitable when the number of graph convolutions tends to infinity, in this paper, we precisely characterize the mechanism behind the phenomenon via a non-asymptotic analysis. Specifically, we distinguish between two different effects when applying graph convolutions-an undesirable mixing effect that homogenizes node representations in different classes, and a desirable denoising effect that homogenizes node representations in the same class. By quantifying these two effects on random graphs sampled from the Contextual Stochastic Block Model (CSBM), we show that oversmoothing happens once the mixing effect starts to dominate the denoising effect, and the number of layers required for this transition is O(log N/ log(log N )) for sufficiently dense graphs with N nodes. We also extend our analysis to study the effects of Personalized PageRank (PPR), or equivalently, the effects of initial residual connections on oversmoothing. Our results suggest that while PPR mitigates oversmoothing at deeper layers, PPR-based architectures still achieve their best performance at a shallow depth and are outperformed by the graph convolution approach on certain graphs. Finally, we support our theoretical results with numerical experiments, which further suggest that the oversmoothing phenomenon observed in practice can be magnified by the difficulty of optimizing deep GNN models.Why does oversmoothing happen at a relatively shallow depth? Can we quantitatively model the effect of applying a finite number of graph convolutions and theoretically predict the "sweet spot" for the choice of depth?In this paper, we propose a non-asymptotic analysis framework to study the effects of graph convolutions and oversmoothing using the Contextual Stochastic Block Model (CSBM)[18]. The CSBM mimics the community structure | A NON-ASYMPTOTIC ANALYSIS OF OVERSMOOTHING IN GRAPH NEURAL NETWORKS |
d235613482 | As a step towards improving the abstract reasoning capability of machines, we aim to solve Raven's Progressive Matrices (RPM) with neural networks, since solving RPM puzzles is highly correlated with human intelligence. Unlike previous methods that use auxiliary annotations or assume hidden rules to produce appropriate feature representation, we only use the ground truth answer of each question for model learning, aiming for an intelligent agent to have a strong learning capability with a small amount of supervision. Based on the RPM problem formulation, the correct answer filled into the missing entry of the third row/column has to best satisfy the same rules shared between the first two rows/columns. Thus we design a simple yet effective Dual-Contrast Network (DCNet) to exploit the inherent structure of RPM puzzles. Specifically, a rule contrast module is designed to compare the latent rules between the filled row/column and the first two rows/columns; a choice contrast module is designed to increase the relative differences between candidate choices. Experimental results on the RAVEN and PGM datasets show that DCNet outperforms the state-of-the-art methods by a large margin of 5.77%. Further experiments on few training samples and model generalization also show the effectiveness of DCNet. Code is available at https://github.com/visiontao/dcnet. Published as a conference paper at ICLR 2021 LEN (Zheng et al., 2019), MXGNet (Wang et al., 2020) and ACL (Kim et al., 2020), not only use the ground truth answer of each RPM question but also the auxiliary annotations (such as logical rules with shape, size, color, number, AND, OR, XOR) to learn the appropriate feature representation. Although auxiliary annotations provide lots of priors about how the logical rules are applied, noticeable performance improvement is not always obtained on different RPM problems, such as in the results reported inTable 1. Moreover, such a learning strategy requires additional supervision. When auxiliary annotations are not available, it will fail to boost the performance. For example, DRT (Zhang et al., 2019a) cannot be applied to PGM dataset(Santoro et al., 2018)for the lack of structure annotations. To overcome the constraint of using auxiliary annotations, a recent method CoPINet (Zhang et al., 2019b) only uses the ground truth answer of each question. Meanwhile, to produce the feature representation of hidden rules, CoPINet assumes there are at most N attributes in each problem, and each of which is subject to the governance of M rules. However, due to N and M being unknown for arbitrary RPM problems, such an assumption is still too strong. | EFFECTIVE ABSTRACT REASONING WITH DUAL-CONTRAST NETWORK |
d249375375 | We study decentralized policy learning in Markov games where we control a single agent to play with nonstationary and possibly adversarial opponents. Our goal is to develop a no-regret online learning algorithm that (i) takes actions based on the local information observed by the agent and (ii) is able to find the best policy in hindsight. For such a problem, the nonstationary state transitions due to the varying opponent pose a significant challenge. In light of a recent hardness result (Liu et al., 2022), we focus on the setting where the opponent's previous policies are revealed to the agent for decision making. With such an information structure, we propose a new algorithm, Decentralized Optimistic hypeRpolicy mIrror deScent (DORIS), which achieves √ Kregret in the context of general function approximation, where K is the number of episodes. Moreover, when all the agents adopt DORIS, we prove that their mixture policy constitutes an approximate coarse correlated equilibrium. In particular, DORIS maintains a hyperpolicy which is a distribution over the policy space. The hyperpolicy is updated via mirror descent, where the update direction is obtained by an optimistic variant of least-squares policy evaluation. Furthermore, to illustrate the power of our method, we apply DORIS to constrained and vector-valued MDPs, which can be formulated as zero-sum Markov games with a fictitious opponent. | Decentralized Optimistic Hyperpolicy Mirror Descent: Provably No-Regret Learning in Markov Games |
d247922261 | Simplicial Embeddings (SEM) are representations learned through self-supervised learning (SSL), wherein a representation is projected into L simplices of V dimensions each using a softmax operation. This procedure conditions the representation onto a constrained space during pre-training and imparts an inductive bias for group sparsity. For downstream classification, we formally prove that the SEM representation leads to better generalization than an unnormalized representation. Furthermore, we empirically demonstrate that SSL methods trained with SEMs have improved generalization on natural image datasets such as CIFAR-100 and ImageNet. Finally, when used in a downstream classification task, we show that SEM features exhibit emergent semantic coherence where small groups of learned features are distinctly predictive of semantically-relevant classes. arXiv:2204.00616v2 [cs.LG] 30 Sep 2022 | SIMPLICIAL EMBEDDINGS IN SELF-SUPERVISED LEARNING AND DOWNSTREAM CLASSIFICATION |
d222133257 | It is well-established that many iterative sparse reconstruction algorithms can be unrolled to yield a learnable neural network for improved empirical performance. A prime example is learned ISTA (LISTA) where weights, step sizes and thresholds are learned from training data. Recently, Analytic LISTA (ALISTA) has been introduced, combining the strong empirical performance of a fully learned approach like LISTA, while retaining theoretical guarantees of classical compressed sensing algorithms and significantly reducing the number of parameters to learn. However, these parameters are trained to work in expectation, often leading to suboptimal reconstruction of individual targets. In this work we therefore introduce Neurally Augmented ALISTA, in which an LSTM network is used to compute step sizes and thresholds individually for each target vector during reconstruction. This adaptive approach is theoretically motivated by revisiting the recovery guarantees of ALISTA. We show that our approach further improves empirical performance in sparse reconstruction, in particular outperforming existing algorithms by an increasing margin as the compression ratio becomes more challenging. * First two authors have equal contribution | Neurally Augmented ALISTA |
d231698498 | Low-precision deep neural network (DNN) training has gained tremendous attention as reducing precision is one of the most effective knobs for boosting DNNs' training time/energy efficiency. In this paper, we attempt to explore low-precision training from a new perspective as inspired by recent findings in understanding DNN training: we conjecture that DNNs' precision might have a similar effect as the learning rate during DNN training, and advocate dynamic precision along the training trajectory for further boosting the time/energy efficiency of DNN training. Specifically, we propose Cyclic Precision Training (CPT) to cyclically vary the precision between two boundary values which can be identified using a simple precision range test within the first few training epochs. Extensive simulations and ablation studies on five datasets and eleven models demonstrate that CPT's effectiveness is consistent across various models/tasks (including classification and language modeling). Furthermore, through experiments and visualization we show that CPT helps to (1) converge to a wider minima with a lower generalization error and (2) reduce training variance which we believe opens up a new design knob for simultaneously improving the optimization and efficiency of DNN training. Our codes are available at: https://github.com/RICE-EIC/CPT. | CPT: EFFICIENT DEEP NEURAL NETWORK TRAINING VIA CYCLIC PRECISION |
d222272463 | We consider the question: how can you sample good negative examples for contrastive learning? We argue that, as with metric learning, learning contrastive representations benefits from hard negative samples (i.e., points that are difficult to distinguish from an anchor point). The key challenge toward using hard negatives is that contrastive methods must remain unsupervised, making it infeasible to adopt existing negative sampling strategies that use label information. In response, we develop a new class of unsupervised methods for selecting hard negative samples where the user can control the amount of hardness. A limiting case of this sampling results in a representation that tightly clusters each class, and pushes different classes as far apart as possible. The proposed method improves downstream performance across multiple modalities, requires only few additional lines of code to implement, and introduces no computational overhead. | Contrastive Learning with Hard Negative Samples |
d259298217 | Out-of-distribution (OOD) generalization is a challenging machine learning problem yet highly desirable in many high-stake applications. Existing methods suffer from overly pessimistic modeling with low generalization confidence. As generalizing to arbitrary test distributions is impossible, we hypothesize that further structure on the topology of distributions is crucial in developing strong OOD resilience. To this end, we propose topology-aware robust optimization (TRO) that seamlessly integrates distributional topology in a principled optimization framework. More specifically, TRO solves two optimization objectives: (1) Topology Learning which explores data manifold to uncover the distributional topology; (2) Learning on Topology which exploits the topology to constrain robust optimization for tightlybounded generalization risks. We theoretically demonstrate the effectiveness of our approach and empirically show that it significantly outperforms the state of the arts in a wide range of tasks including classification, regression, and semantic segmentation. Moreover, we empirically find the data-driven distributional topology is consistent with domain knowledge, enhancing the explainability of our approach. | TOPOLOGY-AWARE ROBUST OPTIMIZATION FOR OUT-OF-DISTRIBUTION GENERALIZATION |
d239768514 | Humans are expert explorers and foragers. Understanding the computational cognitive mechanisms that support this capability can advance the study of the human mind and enable more efficient exploration algorithms. We hypothesize that humans explore new environments by inferring the structure of unobserved spaces through re-use of spatial information collected from previously explored spaces. Taking inspiration from the neuroscience of repeating map fragments and ideas about program induction, we present a novel "Map Induction" framework, which involves the generation of novel map proposals for unseen environments based on compositions of already-seen spaces in a Hierarchical Bayesian framework. The model thus explicitly reasons about unseen spaces through a distribution of strong spatial priors. We introduce a new behavioral Map Induction Task (MIT) that involves foraging for rewards to compare human performance with state-of-the-art existing models and Map Induction. We show that Map Induction better predicts human behavior than the non-inductive baselines. We also show that Map Induction, when used to augment state-of-the-art approximate planning algorithms, improves their performance. Tarr. Do humans integrate routes into a cognitive map? map-versus landmark-based navigation of novel shortcuts. Journal of Experimental Psychology: Learning, Memory, and Cognition, 31(2):195, 2005. Sabine Gillner and Hanspeter A Mallot. Navigation and acquisition of spatial knowledge in a virtual maze. Journal of cognitive neuroscience, 10(4):445-463, 1998. | MAP INDUCTION: COMPOSITIONAL SPATIAL SUBMAP LEARNING FOR EFFICIENT EXPLORATION IN NOVEL ENVIRONMENTS |
d231719359 | For infinitesimal learning rates, stochastic gradient descent (SGD) follows the path of gradient flow on the full batch loss function. However moderately large learning rates can achieve higher test accuracies, and this generalization benefit is not explained by convergence bounds, since the learning rate which maximizes test accuracy is often larger than the learning rate which minimizes training loss. To interpret this phenomenon we prove that for SGD with random shuffling, the mean SGD iterate also stays close to the path of gradient flow if the learning rate is small and finite, but on a modified loss. This modified loss is composed of the original loss function and an implicit regularizer, which penalizes the norms of the minibatch gradients. Under mild assumptions, when the batch size is small the scale of the implicit regularization term is proportional to the ratio of the learning rate to the batch size. We verify empirically that explicitly including the implicit regularizer in the loss can enhance the test accuracy when the learning rate is small.Published as a conference paper at ICLR 2021 therefore helps explain both why finite learning rates can aid generalization, and why SGD can achieve higher test accuracies than GD. We assume that each training example is sampled once per epoch, in line with best practice (Bottou, 2012), and we confirm empirically that explicitly including the implicit regularization term of SGD in the training loss can enhance the test accuracy when the learning rate is small. Furthermore, we prove that if the batch size is small and the gradients are sufficiently diverse, then the expected magnitude of the implicit regularization term of SGD is proportional to the ratio of the learning rate to the batch size(Goyal et al., 2017;.We note that many previous authors have sought to explain the generalization benefit of SGD using an analogy between SGD and stochastic differential equations (SDEs)(Mandt et al., 2017;Jastrzębski et al., 2018;Chaudhari & Soatto, 2018). However this SDE analogy assumes that each minibatch is randomly sampled from the full dataset, which implies that some examples will be sampled multiple times in one epoch. Furthermore, the most common SDE analogy holds only for vanishing learning rates (Yaida, 2019) and therefore misses the generalization benefits of finite learning rates which we identify in this work. An important exception is Li et al.(2017), who applied backward error analysis to identify a modified SDE which holds when the learning rate is finite. However this work still relies on the assumption that minibatches are sampled randomly. It also focused on the convergence rate, and did not discuss the performance of SGD on the test set.Main Result. We now introduce our main result. We define the cost function over parameters ω as C(ω) = (1/N ) | ON THE ORIGIN OF IMPLICIT REGULARIZATION IN STOCHASTIC GRADIENT DESCENT |
d252531851 | Learning to optimize is a rapidly growing area that aims to solve optimization problems or improve existing optimization algorithms using machine learning (ML). In particular, the graph neural network (GNN) is considered a suitable ML model for optimization problems whose variables and constraints are permutation-invariant, for example, the linear program (LP). While the literature has reported encouraging numerical results, this paper establishes the theoretical foundation of applying GNNs to solving LPs. Given any size limit of LPs, we construct a GNN that maps different LPs to different outputs. We show that properly built GNNs can reliably predict feasibility, boundedness, and an optimal solution for each LP in a broad class. Our proofs are based upon the recently-discovered connections between the Weisfeiler-Lehman isomorphism test and the GNN. To validate our results, we train a simple GNN and present its accuracy in mapping LPs to their feasibilities and solutions.Corresponding author: Jialin Liu, jialin.liu@alibaba-inc.com. | ON REPRESENTING LINEAR PROGRAMS BY GRAPH NEURAL NETWORKS |
d257496038 | Deep neural networks have been successful in many reinforcement learning settings. However, compared to human learners they are overly data hungry. To build a sample-efficient world model, we apply a transformer to real-world episodes in an autoregressive manner: not only the compact latent states and the taken actions but also the experienced or predicted rewards are fed into the transformer, so that it can attend flexibly to all three modalities at different time steps. The transformer allows our world model to access previous states directly, instead of viewing them through a compressed recurrent state. By utilizing the Transformer-XL architecture, it is able to learn long-term dependencies while staying computationally efficient. Our transformer-based world model (TWM) generates meaningful, new experience, which is used to train a policy that outperforms previous model-free and model-based reinforcement learning algorithms on the Atari 100k benchmark. Our code is available at https://github.com/jrobine/twm. ∞ t=1 γ t−1 r t , where γ ∈ [0, 1) is the discount factor. Learning in imagination consists of three steps that are repeated iteratively: learning the dynamics, learning a policy, and interacting in the real environment. In this section, we describe our world model and policy, concluding with the training procedure. | TRANSFORMER-BASED WORLD MODELS ARE HAPPY WITH 100K INTERACTIONS |
d246996551 | This paper proposes a novel method for deep learning based on the analytical convolution of multidimensional Gaussian mixtures. In contrast to tensors, these do not suffer from the curse of dimensionality and allow for a compact representation, as data is only stored where details exist. Convolution kernels and data are Gaussian mixtures with unconstrained weights, positions, and covariance matrices. Similar to discrete convolutional networks, each convolution step produces several feature channels, represented by independent Gaussian mixtures. Since traditional transfer functions like ReLUs do not produce Gaussian mixtures, we propose using a fitting of these functions instead. This fitting step also acts as a pooling layer if the number of Gaussian components is reduced appropriately. We demonstrate that networks based on this architecture reach competitive accuracy on Gaussian mixtures fitted to the MNIST and ModelNet data sets. | GAUSSIAN MIXTURE CONVOLUTION NETWORKS |
d3473020 | Localization is the problem of estimating the location of an autonomous agent from an observation and a map of the environment. Traditional methods of localization, which filter the belief based on the observations, are sub-optimal in the number of steps required, as they do not decide the actions taken by the agent. We propose "Active Neural Localizer", a fully differentiable neural network that learns to localize accurately and efficiently. The proposed model incorporates ideas of traditional filtering-based localization methods, by using a structured belief of the state with multiplicative interactions to propagate belief, and combines it with a policy model to localize accurately while minimizing the number of steps required for localization. Active Neural Localizer is trained end-to-end with reinforcement learning. We use a variety of simulation environments for our experiments which include random 2D mazes, random mazes in the Doom game engine and a photo-realistic environment in the Unreal game engine. The results on the 2D environments show the effectiveness of the learned policy in an idealistic setting while results on the 3D environments demonstrate the model's capability of learning the policy and perceptual model jointly from raw-pixel based RGB observations. We also show that a model trained on random textures in the Doom environment generalizes well to a photo-realistic office space environment in the Unreal engine. | ACTIVE NEURAL LOCALIZATION |
d53502621 | Neural network quantization is becoming an industry standard to efficiently deploy deep learning models on hardware platforms, such as CPU, GPU, TPU, and FPGAs. However, we observe that the conventional quantization approaches are vulnerable to adversarial attacks. This paper aims to raise people's awareness about the security of the quantized models, and we designed a novel quantization methodology to jointly optimize the efficiency and robustness of deep learning models. We first conduct an empirical study to show that vanilla quantization suffers more from adversarial attacks. We observe that the inferior robustness comes from the error amplification effect, where the quantization operation further enlarges the distance caused by amplified noise. Then we propose a novel Defensive Quantization (DQ) method by controlling the Lipschitz constant of the network during quantization, such that the magnitude of the adversarial noise remains non-expansive during inference. Extensive experiments on CIFAR-10 and SVHN datasets demonstrate that our new quantization method can defend neural networks against adversarial examples, and even achieves superior robustness than their fullprecision counterparts, while maintaining the same hardware efficiency as vanilla quantization approaches. As a by-product, DQ can also improve the accuracy of quantized models without adversarial attack. | DEFENSIVE QUANTIZATION: WHEN EFFICIENCY MEETS ROBUSTNESS |
d52917627 | Real-life control tasks involve matter of various substances-rigid or soft bodies, liquid, gas-each with distinct physical behaviors. This poses challenges to traditional rigid-body physics engines. Particle-based simulators have been developed to model the dynamics of these complex scenes; however, relying on approximation techniques, their simulation often deviates from real world physics, especially in the long term. In this paper, we propose to learn a particle-based simulator for complex control tasks. Combining learning with particle-based systems brings in two major benefits: first, the learned simulator, just like other particle-based systems, acts widely on objects of different materials; second, the particle-based representation poses strong inductive bias for learning: particles of the same type have the same dynamics within. This enables the model to quickly adapt to new environments of unknown dynamics within a few observations. Using the learned simulator, robots have achieved success in complex manipulation tasks, such as manipulating fluids and deformable foam. The effectiveness of our method has also been demonstrated in the real world. Our study helps lay the foundation for robot learning of dynamic scenes with particle-based representations. | LEARNING PARTICLE DYNAMICS FOR MANIPULATING RIGID BODIES, DEFORMABLE OBJECTS, AND FLUIDS |
d252683860 | We study the problem of deployment efficient reinforcement learning (RL) with linear function approximation under the reward-free exploration setting. This is a well-motivated problem because deploying new policies is costly in real-life RL applications. Under the linear MDP setting with feature dimension d and planning horizon H, we propose a new algorithm that collects at most O( d 2 H 5 ǫ 2 ) trajectories within H deployments to identify ǫ-optimal policy for any (possibly data-dependent) choice of reward functions. To the best of our knowledge, our approach is the first to achieve optimal deployment complexity and optimal d dependence in sample complexity at the same time, even if the reward is known ahead of time. Our novel techniques include an exploration-preserving policy discretization and a generalized G-optimal experiment design, which could be of independent interest. Lastly, we analyze the related problem of regret minimization in low-adaptive RL and provide information-theoretic lower bounds for switching cost and batch complexity. | Near-Optimal Deployment Efficiency in Reward-Free Reinforcement Learning with Linear Function Approximation |
d239616181 | Although DALL·E has shown an impressive ability of composition-based systematic generalization in image generation, it requires the dataset of text-image pairs and the compositionality is provided by the text. In contrast, object-centric representation models like the Slot Attention model learn composable representations without the text prompt. However, unlike DALL·E its ability to systematically generalize for zero-shot generation is significantly limited. In this paper, we propose a simple but novel slot-based autoencoding architecture, called SLATE 1 , for combining the best of both worlds: learning object-centric representations that allows systematic generalization in zero-shot image generation without text. As such, this model can also be seen as an illiterate DALL·E model. Unlike the pixel-mixture decoders of existing object-centric representation models, we propose to use the Image GPT decoder conditioned on the slots for capturing complex interactions among the slots and pixels. In experiments, we show that this simple and easy-to-implement architecture not requiring a text prompt achieves significant improvement in in-distribution and out-of-distribution (zero-shot) image generation and qualitatively comparable or better slot-attention structure than the models based on mixture decoders. https://sites.google.com/view/slate-autoencoder | ILLITERATE DALL-E LEARNS TO COMPOSE |
d252668432 | We show that the representation cost of fully connected neural networks with homogeneous nonlinearities -which describes the implicit bias in function space of networks with L 2 -regularization or with losses such as the cross-entropy -converges as the depth of the network goes to infinity to a notion of rank over nonlinear functions. We then inquire under which conditions the global minima of the loss recover the 'true' rank of the data: we show that for too large depths the global minimum will be approximately rank 1 (underestimating the rank); we then argue that there is a range of depths which grows with the number of datapoints where the true rank is recovered. Finally, we discuss the effect of the rank of a classifier on the topology of the resulting class boundaries and show that autoencoders with optimal nonlinear rank are naturally denoising.Published as a conference paper at ICLR 2023 networks. Furthermore, this bias becomes stronger with depth and in the infinite depth limit L → ∞ the rescaled representation cost R(f )/L converges to the L 0 norm w 0 (the number of non-zero entries in w) in the first case and to the rank Rank(A) in the second.For shallow (L = 2) nonlinear networks with a homogeneous activation, the representation cost also takes the form of a L 1 norm(Bach, 2017;Chizat & Bach, 2020;Ongie et al., 2020), leading to sparsity in the effective number of neurons in the hidden layer of the network.However, the representation cost of deeper networks does not resemble any typical norm (L p or not), though it still leads to some form of sparsity(Jacot et al., 2022b). Despite the absence of explicit formula, we will show that the rescaled representation cost R(f )/L converges to some notion of rank in nonlinear networks as L → ∞, in analogy to infinite depth linear networks.CONTRIBUTIONSWe first introduce two notions of rank: the Jacobian rankfor a linear map A and two bijections ψ and φ), we have Rank J (f ) = Rank BN (f ) = RankA. These two notions of rank satisfy the properties (We then show that in the infinite depth limit L → ∞ the rescaled representation cost of DNNs with a general homogeneous nonlinearity is sandwiched between the Jacobian and Bottleneck ranks:Furthermore lim L→∞ R(f ) satisfies properties (2-4) above. We also conjecture that the limiting representation cost equals its upper bound Rank BN (f ).We then study how this bias towards low-rank functions translates to finite but large depths. We first show that for large depths the rescaled norm of the parameters Ŵ 2 /L at any global minimumŴ is upper bounded by 1 + C N /L for a constant C N which depends on the training points. This implies that the resulting function has approximately rank 1 w.r.t. the Jacobian and Bottleneck ranks. This is however problematic if we are trying to fit a 'true function' f * whose 'true rank' k = Rank BN f * is larger than 1. Thankfully we show that if k > 1 the constant C N explodes as N → ∞, so that the above bound ( Ŵ 2 /L ≤ 1+ C N /L) is relevant only for very large depths when N is large. We show another upper bound Ŵ 2 /L ≤ k + C /L with a constant C independent of N , suggesting the existence of a range of intermediate depths where the network recovers the true rank k.Finally, we discuss how rank recovery affects the topology of decision boundaries in classification and leads autoencoders to naturally be denoising, which we confirm with numerical experiments.RELATED WORKSThe implicit bias of deep homogeneous networks has, to our knowledge, been much less studied than those of either linear networks or shallow nonlinear ones. (Ongie & Willett, 2022) study deep networks with only one nonlinear layer (all others being linear). Similarly(Le & Jegelka, 2022)show a low-rank alignment phenomenon in a network whose last layers are linear.Closer to our setup is the analysis of the representation cost of deep homogeneous networks in (Jacot et al., 2022b), which gives two reformulations for the optimization in the definition of the representation cost, with some implications on the sparsity of the representations, though the infinite depth limit is not studied.A very similar analysis of the sparsity effect of large depth on the global minima of L 2 -regularized networks is given in(Timor et al., 2022), however, they only show how the optimal weight matrices are almost rank 1 (and only on average), while we show low-rank properties of the learned function, as well as the existence of a layer with almost rank 1 hidden representations. | IMPLICIT BIAS OF LARGE DEPTH NETWORKS: A NOTION OF RANK FOR NONLINEAR FUNCTIONS |
d53332150 | There is significant past and ongoing research on optimization methods for deep learning. Yet, perhaps surprisingly, there is no generally agreed-upon protocol for the quantitative and reproducible evaluation of such optimizers. We suggest routines and benchmarks for stochastic optimization, with special focus on the unique aspects of deep learning, such as stochasticity, tunability and generalization. As the primary contribution, we present DEEPOBS, a Python package of deep learning optimization benchmarks. The package addresses key challenges in the quantitative assessment of stochastic optimizers, and automates most steps of benchmarking. The library includes a wide and extensible set of ready-to-use realistic optimization problems, such as training Residual Networks for image classification on IMAGENET or character-level language prediction models, as well as popular classics like MNIST and CIFAR-10. The package also provides realistic baseline results for the most popular optimizers on these test problems, ensuring a fair comparison to the competition when benchmarking new optimizers, and without having to run costly experiments. It comes with output back-ends that directly produce L A T E X code for inclusion in academic publications. It supports TENSORFLOW and is available open source. | DEEPOBS: A DEEP LEARNING OPTIMIZER BENCHMARK SUITE |
d263606004 | Diffusion models have recently been shown to be relevant for high-quality speech generation.Most work has been focused on generating spectrograms, and as such, they further require a subsequent model to convert the spectrogram to a waveform (i.e., a vocoder).This work proposes a diffusion probabilistic end-to-end model for generating a raw speech waveform.The proposed model is autoregressive, generating overlapping frames sequentially, where each frame is conditioned on a portion of the previously generated one.Hence, our model can effectively synthesize an unlimited speech duration while preserving high-fidelity synthesis and temporal coherence.We implemented the proposed model for unconditional and conditional speech generation, where the latter can be driven by an input sequence of phonemes, amplitudes, and pitch values.Working on the waveform directly has some empirical advantages.Specifically, it allows the creation of local acoustic behaviors, like vocal fry, which makes the overall waveform sounds more natural.Furthermore, the proposed diffusion model is stochastic and not deterministic; therefore, each inference generates a slightly different waveform variation, enabling abundance of valid realizations.Experiments show that the proposed model generates speech with superior quality compared with other state-of-the-art neural speech generation systems. | DIFFAR: DENOISING DIFFUSION AUTOREGRESSIVE MODEL FOR RAW SPEECH WAVEFORM GENERATION |
d248177810 | The most significant barrier to the advancement of Neural Architecture Search (NAS) is its demand for large computational resources, which hinders scientifically sound empirical evaluations of NAS methods. Tabular NAS benchmarks have alleviated this problem substantially, making it possible to properly evaluate NAS methods in seconds on commodity machines. However, an unintended consequence of tabular NAS benchmarks has been a focus on extremely small architectural search spaces since their construction relies on exhaustive evaluations of the space. This leads to unrealistic results that do not transfer to larger spaces. To overcome this fundamental limitation, we propose a methodology to create cheap NAS surrogate benchmarks for arbitrary search spaces. We exemplify this approach by creating surrogate NAS benchmarks on the existing tabular NAS-Bench-101 and on two widely used NAS search spaces with up to 10 21 architectures (10 13 times larger than any previous tabular NAS benchmark). We show that surrogate NAS benchmarks can model the true performance of architectures better than tabular benchmarks (at a small fraction of the cost), that they lead to faithful estimates of how well different NAS methods work on the original non-surrogate benchmark, and that they can generate new scientific insight. We open-source all our code and believe that surrogate NAS benchmarks are an indispensable tool to extend scientifically sound work on NAS to large and exciting search spaces. * Equal contribution.Published as a conference paper at ICLR 2022 contain more than 10 18 architectures (Zoph & Le, 2017;Liu et al., 2019;Wu et al., 2019a). This discrepancy can cause results gained on tabular NAS benchmarks to not generalize to realistic search spaces; e.g., promising anytime results of local search on tabular NAS benchmarks were indeed shown to not transfer to realistic search spaces(White et al., 2020b).Making things worse, as discussed in the panel of the most recent NAS workshop at ICLR 2021, to succeed in automatically discovering qualitatively new types of architectures (such as, e.g., Transformers (Vaswani et al., 2017)) the NAS community will have to focus on even more expressive search spaces in the future. To not give up the recent progress in terms of reproducibility that tabular NAS benchmarks have brought, we thus need to develop their equivalent for arbitrary search spaces. That is the goal of this paper.Our contributions. Our main contribution is to introduce the concept of surrogate NAS benchmarks that can be constructed for arbitrary NAS search spaces and allow for the same cheap query interface as tabular NAS benchmarks. We substantiate this contribution as follows:X. Chen, R. Wang, M. Cheng, X. Tang, and C. Hsieh. Drnas: Dirichlet neural architecture search. arXiv preprint arXiv:2006.10355, 2020.P. Chrabaszcz, I. Loshchilov, and F. Hutter. A downsampled variant of imagenet as an alternative to the cifar datasets. arXiv preprint arXiv:1707.08819, 2017. | SURROGATE NAS BENCHMARKS: GOING BEYOND THE LIMITED SEARCH SPACES OF TABULAR NAS BENCHMARKS |
d261494012 | Recent progress in large language models (LLMs) has enabled the processing of long texts consisting of tens of thousands of tokens, which are sufficient for numerous conventional NLP tasks. Many LLMs are trained/fine-tuned to perform zero-shot/few-shot inference using instruction-based prompts. Crafting prompts for these LLMs typically requires the user to provide a detailed task description, examples of context and completion (demonstrations), and single example of context for inference. This regular prompt baseline is referred to as "SinglePrompt" in this paper. However, for NLP tasks where each data point for inference is not necessarily lengthy, the token count for instructions and few-shot examples in the prompt may be considerably larger than that of the data point, resulting in lower token-resource utilization compared with encoder-based models like finetuned BERT. This cost-efficiency issue, affecting inference speed and compute budget, counteracts the many benefits LLMs have to offer. This paper aims to alleviate the preceding problem by batching multiple data points into a single prompt, a prompting strategy we refer to as "BatchPrompt" 1 . This strategy increases the "density" of data points, which in turn leads to improved token utilization. Applying BatchPrompt naïvely, however, is very challenging due to significant performance degradation, as observed in our experiments. We also noticed varying inference outcomes for the same data point appearing in different positions within a prompt. To address the quality issue while remain high tokenresource utilization, we introduce Batch Permutation and Ensembling (BPE) for BatchPrompt, a simple way that recovers labeling quality through majority votes from data points placed in varying positions in a batch at the price of more token usage. To counterbalance the additional token usage caused by the voting process, we further propose Self-reflection-guided EArly Stopping (SEAS), which can terminate the voting process early for data points the LLM confidently handles. Our comprehensive experimental evaluation demonstrates that BPE +SEAS can boost the performance of BatchPrompt with a striking margin on a range of popular NLP tasks, including question answering (Boolq), textual entailment (RTE), and duplicate questions identification (QQP). These performances are even competitive with/higher than single-data prompting (SinglePrompt), while Batch-Prompt requires much fewer LLM calls and input tokens (For SinglePrompt v.s. BatchPrompt+BPE +SEAS with batch size 32, using just 15.7% the number of LLM calls, Boolq accuracy 90.6% → 90.9% with 27.4% tokens, QQP accuracy 87.2% → 88.4% with 18.6% tokens, RTE accuracy 91.5% → 91.1% with 30.8% tokens). We hope our simple yet effective approach will shed light on the future research of large language models. The code will be released. | BATCHPROMPT: ACCOMPLISH MORE WITH LESS |
d227162606 | Continual learning deals with training models on new tasks and datasets in an online fashion. One strand of research has used probabilistic regularization for continual learning, with two of the main approaches in this vein being Online Elastic Weight Consolidation (Online EWC) and Variational Continual Learning (VCL). VCL employs variational inference, which in other settings has been improved empirically by applying likelihood-tempering. We show that applying this modification to VCL recovers Online EWC as a limiting case, allowing for interpolation between the two approaches. We term the general algorithm Generalized VCL (GVCL). In order to mitigate the observed overpruning effect of VI, we take inspiration from a common multi-task architecture, neural networks with task-specific FiLM layers, and find that this addition leads to significant performance gains, specifically for variational methods. In the small-data regime, GVCL strongly outperforms existing baselines. In larger datasets, GVCL with FiLM layers outperforms or is competitive with existing baselines in terms of accuracy, whilst also providing significantly better calibration. | GENERALIZED VARIATIONAL CONTINUAL LEARNING |
d247475741 | We propose a new differentiable probabilistic model over DAGs (DP-DAG). DP-DAG allows fast and differentiable DAG sampling suited to continuous optimization. To this end, DP-DAG samples a DAG by successively (1) sampling a linear ordering of the node and (2) sampling edges consistent with the sampled linear ordering. We further propose VI-DP-DAG, a new method for DAG learning from observational data which combines DP-DAG with variational inference. Hence, VI-DP-DAG approximates the posterior probability over DAG edges given the observed data. VI-DP-DAG is guaranteed to output a valid DAG at any time during training and does not require any complex augmented Lagrangian optimization scheme in contrast to existing differentiable DAG learning approaches. In our extensive experiments, we compare VI-DP-DAG to other differentiable DAG learning baselines on synthetic and real datasets. VI-DP-DAG significantly improves DAG structure and causal mechanism learning while training faster than competitors. | DIFFERENTIABLE DAG SAMPLING |
d232013402 | Graph neural networks have been widely used on modeling graph data, achieving impressive results on node classification and link prediction tasks. Yet, obtaining an accurate representation for a graph further requires a pooling function that maps a set of node representations into a compact form. A simple sum or average over all node representations considers all node features equally without consideration of their task relevance, and any structural dependencies among them. Recently proposed hierarchical graph pooling methods, on the other hand, may yield the same representation for two different graphs that are distinguished by the Weisfeiler-Lehman test, as they suboptimally preserve information from the node features. To tackle these limitations of existing graph pooling methods, we first formulate the graph pooling problem as a multiset encoding problem with auxiliary information about the graph structure, and propose a Graph Multiset Transformer (GMT) which is a multi-head attention based global pooling layer that captures the interaction between nodes according to their structural dependencies. We show that GMT satisfies both injectiveness and permutation invariance, such that it is at most as powerful as the Weisfeiler-Lehman graph isomorphism test. Moreover, our methods can be easily extended to the previous node clustering approaches for hierarchical graph pooling. Our experimental results show that GMT significantly outperforms state-of-the-art graph pooling methods on graph classification benchmarks with high memory and time efficiency, and obtains even larger performance gain on graph reconstruction and generation tasks. | ACCURATE LEARNING OF GRAPH REPRESENTATIONS WITH GRAPH MULTISET POOLING |
d264288929 | We study online reinforcement learning in linear Markov decision processes with adversarial losses and bandit feedback, without prior knowledge on transitions or access to simulators.We introduce two algorithms that achieve improved regret performance compared to existing approaches.The first algorithm, although computationally inefficient, ensures a regret of O( √ K), where K is the number of episodes.This is the first result with the optimal K dependence in the considered setting.The second algorithm, which is based on the policy optimization framework, guarantees a regret of O(K 3 /4 ) and is computationally efficient.Both our results significantly improve over the state-of-the-art: a computationally inefficient algorithm by Kong et al. [2023] with O(K 4 /5 + poly( 1 /λmin)) regret, for some problem-dependent constant λ min that can be arbitrarily close to zero, and a computationally efficient algorithm by Sherman et al. [2023b] with O(K 6 /7 ) regret.* The authors are listed in alphabetical order.is the horizon length.The challenge is that this conversion depends on the transition of the MDP, which is not available to the learner.Therefore, the learner has to estimate the feature of every policy during the learning process.Previous work in this direction [Kong et al., 2023] faced obstacles in controlling the estimation error and was only able to show a K 4 /5 +poly( 1 /λ min ) regret bound assuming there exists an exploratory policy inducing a covariance matrix λ min I.We addressed the obstacles through 1) state space discretization (Section 3.2), and 2) model-free estimation for the occupancy measure of policies over the discretized state space (Section 3.3).These allow us to emulate the success in the tabular case [Jin et al., 2020a] and obtain the tight √ K regret.Efficient K 3 /4 algorithm.The efficient algorithm is based on the policy optimization framework[Luo et al., 2021].Different from previous works that all use exponential weights, we use Follow-the-Regularized-Leader (FTRL) with log-determinant (logdet) barrier regularizer to perform policy updates, which has the benefit of keeping the algorithm more stable [Zimmert and Lattimore, 2022, Liu et al., 2023a].We carefully combine logdet-FTRL with existing algorithmic/analysis techniques to further improve the regret bound.These include 1) an initial exploration phase to control the transition estimation error [Sherman et al., 2023a], 2) optimistic least-square policy evaluation in bonus construction [Sherman et al., 2023b], 3) dilated bonus construction[Luo et al., 2021], and 4) a tighter concentration bound for covariance matrix estimation [Liu et al., 2023a].Related WorkIn this subsection, we review prior works on adversarial MDPs and policy optimization.Learning in Adversarial MDPs.Adversarial MDPs refer to a class of MDP problems where the transition is fixed while the loss function changes over time.Learning adversarial tabular MDPs under bandit feedback and unknown transition has been extensively studied [ | Towards Optimal Regret in Adversarial Linear MDPs with Bandit Feedback |
d258999478 | Recent advancements in neural vocoding are predominantly driven by Generative Adversarial Networks (GANs) operating in the time-domain.While effective, this approach neglects the inductive bias offered by time-frequency representations, resulting in reduntant and computionally-intensive upsampling operations.Fourierbased time-frequency representation is an appealing alternative, aligning more accurately with human auditory perception, and benefitting from well-established fast algorithms for its computation.Nevertheless, direct reconstruction of complexvalued spectrograms has been historically problematic, primarily due to phase recovery issues.This study seeks to close this gap by presenting Vocos, a new model that directly generates Fourier spectral coefficients.Vocos not only matches the state-of-the-art in audio quality, as demonstrated in our evaluations, but it also substantially improves computational efficiency, achieving an order of magnitude increase in speed compared to prevailing time-domain neural vocoding approaches.The source code and model weights have been open-sourced at https://github.com/charactr-platform/vocos. | VOCOS: CLOSING THE GAP BETWEEN TIME-DOMAIN AND FOURIER-BASED NEURAL VOCODERS FOR HIGH-QUALITY AUDIO SYNTHESIS |
d264436566 | GHOST ON THE SHELL: AN EXPRESSIVE REPRESENTATION OF GENERAL 3D SHAPES | |
d219708319 | Robotic manipulation tasks, such as wiping with a soft sponge, require control from multiple rich sensory modalities. Human-robot interaction, aimed at teaching robots, is difficult in this setting as there is potential for mismatch between human and machine comprehension of the rich data streams. We treat the task of interpretable learning from demonstration as an optimisation problem over a probabilistic generative model. To account for the high-dimensionality of the data, a high-capacity neural network is chosen to represent the model. The latent variables in this model are explicitly aligned with high-level notions and concepts that are manifested in a set of demonstrations. We show that such alignment is best achieved through the use of labels from the end user, in an appropriately restricted vocabulary, in contrast to the conventional approach of the designer picking a prior over the latent variables. Our approach is evaluated in the context of a table-top robot manipulation task performed by a PR2 robot -that of dabbing liquids with a sponge (forcefully pressing a sponge and moving it along a surface). | Learning from Demonstration with Weakly Supervised Disentanglement |
d231986303 | Assessing the complexity of functions computed by a neural network helps us understand how the network will learn and generalize. One natural measure of complexity is how the network distorts length -if the network takes a unit-length curve as input, what is the length of the resulting curve of outputs? It has been widely believed that this length grows exponentially in network depth. We prove that in fact this is not the case: the expected length distortion does not grow with depth, and indeed shrinks slightly, for ReLU networks with standard random initialization. We also generalize this result by proving upper bounds both for higher moments of the length distortion and for the distortion of higher-dimensional volumes. These theoretical results are corroborated by our experiments.2. We prove bounds on higher moments of the length distortion, giving upper bounds that hold with high probability (Thm. 4.1). We also obtain similar results for the distortion in the volume of higher-dimensional manifolds of inputs (Thm. 4.2).3. We empirically verify that our theoretical results accurately predict observed behavior for networks at initialization, while previous bounds are loose and fail to capture subtle architecture dependencies.It is worth explaining why our conclusions differ from those of[22,23]. First, prior authors prove only lower bounds on the expected length distortion, while we use different methodology to calculate * Equal contribution | Deep ReLU Networks Preserve Expected Length |
d259937490 | Diffusion models have recently emerged as powerful generative priors for solving inverse problems.However, training diffusion models in the pixel space are both data-intensive and computationally demanding, which restricts their applicability as priors for high-dimensional real-world data such as medical images.Latent diffusion models, which operate in a much lower-dimensional space, offer a solution to these challenges.However, incorporating latent diffusion models to solve inverse problems remains a challenging problem due to the nonlinearity of the encoder and decoder.To address these issues, we propose ReSample, an algorithm that can solve general inverse problems with pre-trained latent diffusion models.Our algorithm incorporates data consistency by solving an optimization problem during the reverse sampling process, a concept that we term as hard data consistency.Upon solving this optimization problem, we propose a novel resampling scheme to map the measurement-consistent sample back onto the noisy data manifold and theoretically demonstrate its benefits.Lastly, we apply our algorithm to solve a wide range of linear and nonlinear inverse problems in both natural and medical images, demonstrating that our approach outperforms existing state-of-the-art approaches, including those based on pixel-space diffusion models. | Solving Inverse Problems with Latent Diffusion Models via Hard Data Consistency |
d252873172 | Without the demand of training in reality, humans can easily detect a known concept simply based on its language description. Empowering deep learning with this ability undoubtedly enables the neural network to handle complex vision tasks, e.g., object detection, without collecting and annotating real images. To this end, this paper introduces a novel challenging learning paradigm Imaginary-Supervised Object Detection (ISOD), where neither real images nor manual annotations are allowed for training object detectors. To resolve this challenge, we propose IMAGINARYNET, a framework to synthesize images by combining pretrained language model and text-to-image synthesis model. Given a class label, the language model is used to generate a full description of a scene with a target object, and the text-to-image model deployed to generate a photo-realistic image. With the synthesized images and class labels, weakly supervised object detection can then be leveraged to accomplish ISOD. By gradually introducing real images and manual annotations, IMAGINARYNET can collaborate with other supervision settings to further boost detection performance. Experiments show that IMAGINARYNET can (i) obtain about 70% performance in ISOD compared with the weakly supervised counterpart of the same backbone trained on real data, (ii) significantly improve the baseline while achieving state-of-the-art or comparable performance by incorporating IMAGINARYNET with other supervision settings. Our code is publicly available at https://github.com/kodenii/ImaginaryNet. | IMAGINARYNET: LEARNING OBJECT DETECTORS WITHOUT REAL IMAGES AND ANNOTATIONS |
d220936168 | We study the global convergence and global optimality of actor-critic, one of the most popular families of reinforcement learning algorithms. While most existing works on actor-critic employ bi-level or two-timescale updates, we focus on the more practical single-timescale setting, where the actor and critic are updated simultaneously. Specifically, in each iteration, the critic update is obtained by applying the Bellman evaluation operator only once while the actor is updated in the policy gradient direction computed using the critic. Moreover, we consider two function approximation settings where both the actor and critic are represented by linear or deep neural networks. For both cases, we prove that the actor sequence converges to a globally optimal policy at a sublinear O(K −1/2 ) rate, where K is the number of iterations. To the best of our knowledge, we establish the rate of convergence and global optimality of single-timescale actor-critic with linear function approximation for the first time. Moreover, under the broader scope of policy optimization with nonlinear function approximation, we prove that actor-critic with deep neural network finds the globally optimal policy at a sublinear rate for the first time.the bi-level setting or the two-timescale setting, which are seldom adopted in practice. In particular, under the bi-level setting (Yang et al., 2019a; Agarwal et al., 2019; Fu et al., 2019; Liu et al., 2019; Abbasi-Yadkori et al., 2019a,b;Cai et al., 2019;Hao et al., 2020; Mei et al., 2020;Bhandari and Russo, 2020), the actor is updated only after the critic solves the policy evaluation sub-problem completely, which is equivalent to applying the Bellman evaluation operator to the previous critic for infinite times. Consequently, actor-critic under the bi-level setting is a double-loop iterative algorithm where the inner loop is allocated for solving the policy evaluation sub-problem of the critic. In terms of theoretical analysis, such a double-loop structure decouples the analysis for the actor and critic. For the actor, the problem is essentially reduced to analyzing the convergence of a variant of the policy gradient method (Sutton et al., 2000; Kakade, 2002) where the error of the gradient estimate depends on the policy evaluation error of the critic. Besides, under the two-timescale setting(Borkar and Konda, 1997;Konda and Tsitsiklis, 2000; Xu et al., 2020; Wu et al., 2020; Hong et al., 2020), the actor and the critic are updated simultaneously, but with disparate stepsizes. More concretely, the stepsize of the actor is set to be much smaller than that of the critic, with the ratio between these stepsizes converging to zero. In an asymptotic sense, such a separation between stepsizes ensures that the critic completely solves its policy evaluation sub-problem asymptotically. In other words, such a two-timescale scheme results in a separation between actor and critic in an asymptotic sense, which leads to asymptotically unbiased policy gradient estimates. In sum, in terms of convergence analysis, the existing theory of actor-critic hinges on decoupling the analysis for critic and actor, which is ensured via focusing on the bi-level or two-timescale settings.However, most practical implementations of actor-critic are under the single-timescale setting (Peters and Schaal, 2008a; Schulman et al., 2015; Mnih et al., 2016; Schulman et al., 2017; Haarnoja et al., 2018), where the actor and critic are simultaneously updated, and particularly, the actor is updated without the critic reaching an approximate solution to the policy evaluation sub-problem. Meanwhile, in comparison with the twotimescale setting, the actor is equipped with a much larger stepsize in the the single-timescale setting such that the asymptotic separation between the analysis of actor and critic is no longer valid. Furthermore, when it comes to function approximation, most existing works only analyze the convergence of actor-critic with either linear function approximation (Xu et al., 2020; Wu et al., 2020; Hong et al., 2020), or shallow-neural-network parameterization (Wang et al., 2019; Liu et al., 2019). In contrast, practically used actor-critic methods such as asynchronous advantage actor-critic (Mnih et al., 2016) and soft actorcritic (Haarnoja et al., 2018) oftentimes represent both the actor and critic using deep neural networks. Thus, the following question is left open:Does single-timescale actor-critic provably find a globally optimal policy under the function approximation setting, especially when deep neural networks are employed?To answer such a question, we make the first attempt to investigate the convergence and global optimality of single-timescale actor-critic with linear and neural network function approximation. In particular, we focus on the family of energy-based policies and aim to find the optimal policy within this class. Here we represent both the energy function and the critic as linear or deep neural network functions. In our actorcritic algorithm, the actor update follows proximal policy optimization (PPO) (Schulman et al., 2017) and the critic update is obtained by applying the Bellman evaluation operator only once to the current critic iterate. As a result, the actor is updated before the critic solves the policy evaluation sub-problem. Such a coupled updating structure persists even when the number of iterations goes to infinity, which implies that 2 the update direction of the actor is always biased compared with the policy gradient direction. This brings an additional challenge that is absent in the bi-level and the two-timescale settings, where the actor and critic are decoupled asymptotically.To tackle such a challenge, our analysis captures the joint effect of actor and critic updates on the objective function, dubbed as the "double contraction" phenomenon, which plays a pivotal role for the success of single-timescale actor-critic. Specifically, thanks to the discount factor of the MDP, the Bellman evaluation operator is contractive, which implies that, after each update, the critic makes noticeable progress by moving towards the value function associated with the current actor. As a result, although we use a biased estimate of the policy gradient, thanks to the contraction brought by the discount factor, the accumulative effect of the biases is controlled. Such a phenomenon enables us to characterize the progress of each iteration of joint actor and critic update, and thus yields the convergence to the globally optimal policy. In particular, for both the linear and neural settings, we prove that, single-timescale actor-critic finds a O(K −1/2 )-globally optimal policy after K iterations. To the best of our knowledge, we seem to establish the first theoretical guarantee of global convergence and global optimality for actor-critic with function approximation in the single-timescale setting. Moreover, under the broader scope of policy optimization with nonlinear function approximation, our work seems to prove convergence and optimality guarantees for actor-critic with deep neural network for the first time.Contribution. Our contribution is two-fold. First, in the single-timescale setting with linear function approximation, we prove that, after K iterations of actor and critic updates, actor-critic returns a policy that is at most O(K −1/2 ) inferior to the globally optimal policy. Second, when both the actor and critic are represented by deep neural networks, we prove a similar O(K −1/2 ) rate of convergence to the globally optimal policy when the architecture of the neural networks are properly chosen. Related Work. Our work extends the line of works on the convergence of actor-critic under the function approximation setting. In particular, actor-critic is first introduced in Sutton et al. (2000); Konda and Tsitsiklis (2000). Later, Kakade (2002); Peters and Schaal (2008b) propose the natural actor-critic method which updates the policy via the natural gradient (Amari, 1998) direction. The convergence of (natural) actorcritic with linear function approximation are studied in Bhatnagar et al. (2008, 2009); Bhatnagar (2010); Castro and Meir (2010); Maei (2018). However, these works only characterize the asymptotic convergence of actor-critic and their proofs all resort to tools from stochastic approximation via ordinary differential equations (Borkar, 2008). As a result, these works only show that actor-critic with linear function approximation converges to the set of stable equilibria of a set of ordinary differential equations. Recently, Zhang et al. (2019) propose a variant of actor-critic where Monte-Carlo sampling is used to ensure the critic and the policy gradient estimates are unbiased. Although they incorporate nonlinear function approximation in the actor, they only establish finite-time convergence result to a stationary point of the expected total reward. Moreover, due to having an inner loop for solving the policy evaluation sub-problem, they focus on the bi-level setting. Moreover, under the two-timescale setting, Wu et al. (2020); Xu et al. (2020) show that actor-critic with linear function approximation finds an ε-stationary point with O(ε −5/2 ) samples, where ε measures the squared norm of the policy gradient. All of these results establish the convergence of actor-critic, without characterizing the optimality of the policy obtained by actor-critic. In terms of the global optimality of actor-critic, Fazel et al. (2018); Malik et al. (2018); Tu and Recht (2018); Yang et al. (2019a); Bu et al. (2019); Fu et al. (2019) show that policy gradient and bi-level actor-3 critic methods converge to the globally optimal policies under the linear-quadratic setting, where the state transitions follow a linear dynamical system and the reward function is quadratic. For general MDPs, Bhandari and Russo (2019) recently prove the global optimality of vanilla policy gradient under the assumption that the families of policies and value functions are both convex. In addition, our work is also related to Liu et al. (2019) and Wang et al. (2019), where they establish the global optimality of proximal policy optimization and (natural) actor-critic, respectively, where both the actor and critic are parameterized by two-layer neural networks. Our work is also related to Agarwal et al. (2019); Abbasi-Yadkori et al. (2019a,b); Cai et al. (2019); Hao et al. (2020); Mei et al. (2020); Bhandari and Russo(2020), which focus on characterizing the optimality of natural policy gradient in tabular and/or linear settings. However, these aforementioned works all focus on bi-level actor-critic, where the actor is updated only after the critic solves the policy evaluation sub-problem to an approximate optimum. Besides, these works consider linear or two-layer neural network function approximations whereas we focus on the setting with deep neural networks. Furthermore, under the two-timescale setting, Xu et al. (2020); Hong et al. (2020) prove that linear actor-critic requires a sample complexity of O(ε −4 ) for obtaining an ε-globally optimal policy. In comparison, our O(K −1/2 ) convergence for single-timescale actor-critic can be translated into a similar O(ε −4 ) sample complexity directly.Moreover, when reusing the data, our result leads to an improved O(ε −2 ) sample complexity. In addition, our work is also related to Geist et al.(2019), which proposes a variant of policy iteration algorithm with Bregman divergence regularization. Without considering an explicit form of function approximation, their algorithm is shown to converge to the globally optimal policy at a similar O(K −1/2 ) rate, where K is the number of policy updates. In contrast, our method is single-timescale actor-critic with linear or deep neural network function approximation, which enjoys both global convergence and global optimality. Meanwhile, our proof is based on a finite-sample analysis, which involves dealing with the algorithmic errors that track the performance of actor and critic updates as well as the statistical error due to having finite data.Our work is also related to the literature on deep neural networks. Previous works (Daniely, 2017; Jacot et al., ; Gao et al., 2019) analyze the computational and statistical rates of supervised learning methods with overparameterized neural networks. In contrast, our work employs overparameterized deep neural networks in actor-critic for solving RL tasks, which is significantly more challenging than supervised learning due to the interplay between the actor and the critic.Roadmap. In §2, we introduce the background of discounted MDP and actor-critic method. Then in §3, we introduce the two actor-critic methods, where the actors and critics are parameterized using linear functions and deep neural networks. The theoretical results are presented in §4. | Single-Timescale Actor-Critic Provably Finds Globally Optimal Policy |
d258557287 | Diffusion models have emerged as a key pillar of foundation models in visual domains. One of their critical applications is to universally solve different downstream inverse tasks via a single diffusion prior without re-training for each task. Most inverse tasks can be formulated as inferring a posterior distribution over data (e.g., a full image) given a measurement (e.g., a masked image). This is however challenging in diffusion models since the nonlinear and iterative nature of the diffusion process renders the posterior intractable. To cope with this challenge, we propose a variational approach that by design seeks to approximate the true posterior distribution. We show that our approach naturally leads to regularization by denoising diffusion process (RED-diff) where denoisers at different timesteps concurrently impose different structural constraints over the image. To gauge the contribution of denoisers from different timesteps, we propose a weighting mechanism based on signal-to-noise-ratio (SNR). Our approach provides a new variational perspective for solving inverse problems with diffusion models, allowing us to formulate sampling as stochastic optimization, where one can simply apply off-the-shelf solvers with lightweight iterates. Our experiments for various linear and nonlinear image restoration tasks demonstrate the strengths of our method compared with state-of-the-art sampling-based diffusion models. The code can be found at GitHub. | A VARIATIONAL PERSPECTIVE ON SOLVING INVERSE PROBLEMS WITH DIFFUSION MODELS |
d3463660 | We build a virtual agent for learning language in a 2D maze-like world. The agent sees images of the surrounding environment, listens to a virtual teacher, and takes actions to receive rewards. It interactively learns the teacher's language from scratch based on two language use cases: sentence-directed navigation and question answering. It learns simultaneously the visual representations of the world, the language, and the action control. By disentangling language grounding from other computational routines and sharing a concept detection function between language grounding and prediction, the agent reliably interpolates and extrapolates to interpret sentences that contain new word combinations or new words missing from training sentences. The new words are transferred from the answers of language prediction. Such a language ability is trained and evaluated on a population of over 1.6 million distinct sentences consisting of 119 object words, 8 color words, 9 spatial-relation words, and 50 grammatical words. The proposed model significantly outperforms five comparison methods for interpreting zero-shot sentences. In addition, we demonstrate human-interpretable intermediate outputs of the model in the appendix. . A roadmap towards machine intelligence. arXiv preprint arXiv:1511.08130, 2015.Dipendra Misra, John Langford, and Yoav Artzi. Mapping instructions and visual observations to actions with reinforcement learning. In EMNLP, 2017. | INTERACTIVE GROUNDED LANGUAGE ACQUISITION AND GENERALIZATION IN A 2D WORLD |
d211296452 | We consider the task of answering complex multi-hop questions using a corpus as a virtual knowledge base (KB). In particular, we describe a neural module, DrKIT, that traverses textual data like a KB, softly following paths of relations between mentions of entities in the corpus. At each step the module uses a combination of sparse-matrix TFIDF indices and a maximum inner product search (MIPS) on a special index of contextual representations of the mentions. This module is differentiable, so the full system can be trained end-to-end using gradient based methods, starting from natural language inputs. We also describe a pretraining scheme for the contextual representation encoder by generating hard negative examples using existing knowledge bases. We show that DrKIT improves accuracy by 9 points on 3-hop questions in the MetaQA dataset, cutting the gap between text-based and KB-based state-of-the-art by 70%. On HotpotQA, DrKIT leads to a 10% improvement over a BERT-based re-ranking approach to retrieving the relevant passages required to answer a question. DrKIT is also very efficient, processing 10-100x more queries per second than existing multi-hop systems. 1 * Part of this work was done during an internship at Google. 1 | DIFFERENTIABLE REASONING OVER A VIRTUAL KNOWLEDGE BASE |
d246823969 | 1-bit gradient compression and local steps are two representative techniques that enable drastic communication reduction in distributed SGD. Their benefits, however, remain an open question on Adambased large model pre-training (e.g. BERT and GPT). In this paper, we demonstrate the non-linearity in Adam causes slow convergence even when 1-bit compression or local steps are individually applied. To alleviate this limitation, we propose 0/1 Adam that linearizes each Adam step via approximating its optimizer states using their stale estimates and linear correlation. 0/1 Adam performs an Adam-like step to preserve the adaptivity, while its linearity allows utilizing 1-bit compression and local steps simultaneously for wall-clock time speed up. We provide convergence guarantee for 0/1 Adam on smooth non-convex objectives. On various large-scale benchmarks such as BERT-Base, BERT-Large, GPT-2 pre-training and ImageNet, we demonstrate on up to 128 GPUs that 0/1 Adam is able to reduce up to 87% of data volume, 54% of communication rounds, and achieve up to 2× higher training throughput and end-to-end training time reduction compared to the state-of-the-art baseline 1-bit Adam; while enjoying the same statistical convergence speed and end task model accuracy on GLUE dataset and ImageNet validation set. * Corresponds to: yl2967@cornell.edu. arXiv:2202.06009v3 [cs.LG] 22 May 2022 precision communication (full-precision stage 1 ); and then switch to 1 bit when the variance becomes stable (compression stage). While this paradigm avoids compressing non-linear information with a one-time frozen variance, the experimental results from[11]indicate the full-precision stage still incurs non-trivial overhead. Furthermore, 1-bit Adam is restricted in the scope of gradient compression, and cannot be trivially adapted when other techniques are used, such as local steps. Besides, the empirical success of[11]was not substantiated on generative models (GPT-style models), for instance, 175B GPT-3 [3], 530B MT-NLG [4], etc.In this paper, we address this gap by proposing 0/1 Adam. 0/1 Adam breaks the barrier of non-linearity from two aspects: first it adaptively freezes variance, so that given agreement on a stale variance state, the parallel workers only need to communicate momentum that is linearly dependent on the model update; This technique allows reducing the previous two-stage compression scheme to a unified single stage; 2) it leverages the insight that in adjacent Adam steps, the changes to optimizer states are generally bounded, so that with frozen variance, parallel workers can linearly approximate momentum and parameter updates locally without additional synchronization. This further pushes the limit of communication reduction towards its extreme, achieving the state-of-the-art speed up for large-scale model training. To summarize, our contributions are as follows:• We propose 0/1 Adam, a novel optimization method that addresses the non-linearity challenges in Adam when applying aggressive 1-bit quantization and local steps (Section 4). • We provide convergence guarantee of 0/1 Adam on smooth and non-convex objectives (Section 5).• We conduct experiments on a wide range of large-scale model training tasks, including BERT-Base, BERT-Large, GPT-2 pre-training and ImageNet. We demonstrate on up to 128 GPUs that 0/1 Adam is able to reduce up to 87% of data volume, 54% of communication rounds, and achieve up to 2× higher throughput and training time reduction compared to the state-of-the-art 1-bit Adam without compromising end-to-end model accuracy (Section 6). • The 0/1 Adam optimizer and corresponding experimental scripts (e.g. BERT pre-training and GLUE finetuning) have been open sourced in a deep learning optimization library called DeepSpeed 2 . | Maximizing Communication Efficiency for Large-scale Training via 0/1 Adam |
d53732884 | A growing number of learning methods are actually differentiable games whose players optimise multiple, interdependent objectives in parallel -from GANs and intrinsic curiosity to multi-agent RL. Opponent shaping is a powerful approach to improve learning dynamics in these games, accounting for player influence on others' updates. Learning with Opponent-Learning Awareness (LOLA) is a recent algorithm that exploits this response and leads to cooperation in settings like the Iterated Prisoner's Dilemma. Although experimentally successful, we show that LOLA agents can exhibit 'arrogant' behaviour directly at odds with convergence. In fact, remarkably few algorithms have theoretical guarantees applying across all (n-player, non-convex) games. In this paper we present Stable Opponent Shaping (SOS), a new method that interpolates between LOLA and a stable variant named LookAhead. We prove that LookAhead converges locally to equilibria and avoids strict saddles in all differentiable games. SOS inherits these essential guarantees, while also shaping the learning of opponents and consistently either matching or outperforming LOLA experimentally. | STABLE OPPONENT SHAPING IN DIFFERENTIABLE GAMES |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.