Transformer, originally devised for natural language processing, has also attested significant success in computer vision. Thanks to its super expressive power, researchers are investigating ways to deploy transformers to reinforcement learning (RL) and the transformer-based models have manifested their potential in representative RL benchmarks. In this paper, we collect and dissect recent advances on transforming RL by transformer (transformer-based RL or TRL), in order to explore its development trajectory and future trend. We group existing developments in two categories: architecture enhancement and trajectory optimization, and examine the main applications of TRL in robotic manipulation, text-based games, navigation and autonomous driving. For architecture enhancement, these methods consider how to apply the powerful transformer structure to RL problems under the traditional RL framework, which model agents and environments much more precisely than deep RL methods, but they are still limited by the inherent defects of traditional RL algorithms, such as bootstrapping and "deadly triad". For trajectory optimization, these methods treat RL problems as sequence modeling and train a joint state-action model over entire trajectories under the behavior cloning framework, which are able to extract policies from static datasets and fully use the long-sequence modeling capability of the transformer. Given these advancements, extensions and challenges in TRL are reviewed and proposals about future direction are discussed. We hope that this survey can provide a detailed introduction to TRL and motivate future research in this rapidly developing field.
translated by 谷歌翻译
We introduce INSTRUCTOR, a new method for computing text embeddings given task instructions: every text input is embedded together with instructions explaining the use case (e.g., task and domain descriptions). Unlike encoders from prior work that are more specialized, INSTRUCTOR is a single embedder that can generate text embeddings tailored to different downstream tasks and domains, without any further training. We first annotate instructions for 330 diverse tasks and train INSTRUCTOR on this multitask mixture with a contrastive loss. We evaluate INSTRUCTOR on 70 embedding evaluation tasks (66 of which are unseen during training), ranging from classification and information retrieval to semantic textual similarity and text generation evaluation. INSTRUCTOR, while having an order of magnitude fewer parameters than the previous best model, achieves state-of-the-art performance, with an average improvement of 3.4% compared to the previous best results on the 70 diverse datasets. Our analysis suggests that INSTRUCTOR is robust to changes in instructions, and that instruction finetuning mitigates the challenge of training a single model on diverse datasets.
translated by 谷歌翻译
We propose eXtensible Prompt (X-Prompt) for prompting a large language model (LLM) beyond natural language (NL). X-Prompt instructs an LLM with not only NL but also an extensible vocabulary of imaginary words that are introduced to help represent what NL words hardly describe, allowing a prompt to be more descriptive. Like NL prompts, X-Prompt is out-of-distribution (OOD) robust, for which we propose context-guided learning with prompt augmentation to learn its imaginary words for general usability, enabling them to use in different prompt contexts for fine-grain specifications. The promising results of X-Prompt demonstrate its potential of approaching advanced interaction between humans and LLMs to bridge their communication gap.
translated by 谷歌翻译
The recently developed discrete diffusion models perform extraordinarily well in the text-to-image task, showing significant promise for handling the multi-modality signals. In this work, we harness these traits and present a unified multimodal generation model that can conduct both the "modality translation" and "multi-modality generation" tasks using a single model, performing text-based, image-based, and even vision-language simultaneous generation. Specifically, we unify the discrete diffusion process for multimodal signals by proposing a unified transition matrix. Moreover, we design a mutual attention module with fused embedding layer and a unified objective function to emphasise the inter-modal linkages, which are vital for multi-modality generation. Extensive experiments indicate that our proposed method can perform comparably to the state-of-the-art solutions in various generation tasks.
translated by 谷歌翻译
在各种图像处理和计算机视觉任务中经常遇到颜色图像Denoising。一种传统的策略是将RGB图像转换为较小相关的颜色空间,并分别将新空间的每个通道定义。但是,这种策略无法完全利用渠道之间的相关信息,并且不足以获得令人满意的结果。为了解决这个问题,本文提出了一个新的多通道优化模型,用于在核定标准下减去Frobenius规范最小化框架下的颜色图像Deno。具体而言,基于块匹配,将颜色图像分解为重叠的RGB补丁。对于每个补丁,我们堆叠其相似的邻居以形成相应的补丁矩阵。提出的模型是在补丁矩阵上执行的,以恢复其无噪声版本。在恢复过程中,a)引入权重矩阵以充分利用通道之间的噪声差; b)单数值是自适应缩小的,而无需分配权重。有了他们,提议的模型可以在保持简单的同时取得有希望的结果。为了解决提出的模型,基于乘数框架的交替方向方法构建了准确有效的算法。每个更新步骤的解决方案可以在封闭式中分析表达。严格的理论分析证明了所提出的算法产生的解决方案序列会收敛到其各自的固定点。合成和真实噪声数据集的实验结果证明了所提出的模型优于最先进的模型。
translated by 谷歌翻译
很少有视觉识别是指从一些标记实例中识别新颖的视觉概念。通过将查询表示形式与类表征进行比较以预测查询实例的类别,许多少数射击的视觉识别方法采用了基于公制的元学习范式。但是,当前基于度量的方法通常平等地对待所有实例,因此通常会获得有偏见的类表示,考虑到并非所有实例在总结了类级表示的实例级表示时都同样重要。例如,某些实例可能包含无代表性的信息,例如过多的背景和无关概念的信息,这使结果偏差。为了解决上述问题,我们提出了一个新型的基于公制的元学习框架,称为实例自适应类别表示网络(ICRL-net),以进行几次视觉识别。具体而言,我们开发了一个自适应实例重新平衡网络,具有在生成班级表示,通过学习和分配自适应权重的不同实例中的自适应权重时,根据其在相应类的支持集中的相对意义来解决偏见的表示问题。此外,我们设计了改进的双线性实例表示,并结合了两个新型的结构损失,即,阶层内实例聚类损失和阶层间表示区分损失,以进一步调节实例重估过程并完善类表示。我们对四个通常采用的几个基准测试:Miniimagenet,Tieredimagenet,Cifar-FS和FC100数据集进行了广泛的实验。与最先进的方法相比,实验结果证明了我们的ICRL-NET的优势。
translated by 谷歌翻译
大型语言模型(LLM)已在一系列自然语言理解任务上实现了最先进的表现。但是,这些LLM可能依靠数据集偏差和文物作为预测的快捷方式。这极大地损害了他们的分布(OOD)概括和对抗性鲁棒性。在本文中,我们对最新发展的综述,这些发展解决了LLMS的鲁棒性挑战。我们首先介绍LLM的概念和鲁棒性挑战。然后,我们介绍了在LLM中识别快捷方式学习行为的方法,表征了快捷方式学习的原因以及引入缓解解决方案。最后,我们确定了关键挑战,并将这一研究线的联系引入其他方向。
translated by 谷歌翻译
尽管参数有效调整(PET)方法在自然语言处理(NLP)任务上显示出巨大的潜力,但其有效性仍然对计算机视觉(CV)任务的大规模转向进行了研究。本文提出了Conv-Adapter,这是一种专为CONCNET设计的PET模块。 Conv-Adapter具有轻巧的,可转让的域和架构,不合时宜,并且在不同的任务上具有广义性能。当转移下游任务时,Conv-Adapter将特定于任务的特征调制到主链的中间表示,同时保持预先训练的参数冻结。通过仅引入少量可学习的参数,例如,仅3.5%的RESNET50的完整微调参数,Conv-Adapter优于先前的宠物基线方法,并实现可比性或超过23个分类任务的全面调查的性能。它还在几乎没有分类的情况下表现出卓越的性能,平均利润率为3.39%。除分类外,Conv-Adapter可以推广到检测和细分任务,其参数降低了50%以上,但性能与传统的完整微调相当。
translated by 谷歌翻译
在线二手匹配是在线算法中的一个基本问题。目的是匹配两组顶点,以最大化边缘权重的总和,在该顶点中,对于一组顶点,每个顶点及其相应的边缘重量以序列形式出现。当前,在实际的建议系统或搜索引擎中,权重是由用户的深度表示与项目深度表示之间的内部产品决定的。标准的在线匹配需要支付$ nd $的时间来线性扫描所有$ n $项目,计算重量(假设每个表示向量都有长度$ d $),然后根据权重决定匹配。但是,实际上,$ n $可能很大,例如在在线电子商务平台中。因此,改善计算权重的时间是一个实践意义的问题。在这项工作中,我们为大约计算权重的理论基础提供了基础。我们表明,借助我们提出的随机数据结构,可以在额定时间内计算权重,同时仍保留匹配算法的竞争比率。
translated by 谷歌翻译
由于联邦学习(FL)的分布性质,研究人员发现FL容易受到后门攻击的影响,该攻击旨在将子任务注入FL而不破坏主要任务的性能。当在FL模型收敛上注入时,单发后门攻击在主要任务和后门子任务上都可以达到高度精度。但是,早期注射的单发后门攻击是无效的,因为:(1)由于正常局部更新的稀释效果,在注射时未达到最大的后门效果; (2)后门效应迅速下降,因为后门将被新的普通本地更新所覆盖。在本文中,我们利用FL模型信息泄漏加强了早期注射的单发后门攻击。我们表明,如果客户在模拟整个人群的分布和梯度的数据集上进行训练,则可以加快FL收敛速度。基于这一观察结果,我们提出了两阶段的后门攻击,其中包括随后的后门攻击的初步阶段。在初步阶段,受攻击者控制的客户首先启动了整个人口分布推理攻击,然后在本地制作的数据集上进行训练,该数据集与梯度和推断分布保持一致。从初步阶段中受益,后来注射的后门实现了更好的有效性,因为后门效应不太可能被普通模型更新稀释。在各种数据异质性设置下,在MNIST数据集上进行了广泛的实验,以评估拟议的后门攻击的有效性。结果表明,即使有防御机制,该提议的后门以成功率和寿命都优于现有的后门攻击。
translated by 谷歌翻译