Empirical studies have identified a range of learnability biases and limitations of transformers, such as a persistent difficulty in learning to compute simple formal languages such as PARITY, and a bias towards low-degree functions. However, theoretical understanding remains limited, with existing expressiveness theory either overpredicting or underpredicting realistic learning abilities. We prove that, under the transformer architecture, the loss landscape is constrained by the input-space sensitivity: Transformers whose output is sensitive to many parts of the input string inhabit isolated points in parameter space, leading to a low-sensitivity bias in generalization. We show theoretically and empirically that this theory unifies a broad array of empirical observations about the learning abilities and biases of transformers, such as their generalization bias towards low sensitivity and low degree, and difficulty in length generalization for PARITY. This shows that understanding transformers' inductive biases requires studying not just their in-principle expressivity, but also their loss landscape.
相關內容
Multilingual pretraining and fine-tuning have remarkably succeeded in various natural language processing tasks. Transferring representations from one language to another is especially crucial for cross-lingual learning. One can expect machine translation objectives to be well suited to fostering such capabilities, as they involve the explicit alignment of semantically equivalent sentences from different languages. This paper investigates the potential benefits of employing machine translation as a continued training objective to enhance language representation learning, bridging multilingual pretraining and cross-lingual applications. We study this question through two lenses: a quantitative evaluation of the performance of existing models and an analysis of their latent representations. Our results show that, contrary to expectations, machine translation as the continued training fails to enhance cross-lingual representation learning in multiple cross-lingual natural language understanding tasks. We conclude that explicit sentence-level alignment in the cross-lingual scenario is detrimental to cross-lingual transfer pretraining, which has important implications for future cross-lingual transfer studies. We furthermore provide evidence through similarity measures and investigation of parameters that this lack of positive influence is due to output separability -- which we argue is of use for machine translation but detrimental elsewhere.
We investigate the behavior of methods that use linear projections to remove information about a concept from a language representation, and we consider the question of what happens to a dataset transformed by such a method. A theoretical analysis and experiments on real-world and synthetic data show that these methods inject strong statistical dependencies into the transformed datasets. After applying such a method, the representation space is highly structured: in the transformed space, an instance tends to be located near instances of the opposite label. As a consequence, the original labeling can in some cases be reconstructed by applying an anti-clustering method.
There is a growing concern that generative AI models will generate outputs closely resembling the copyrighted materials for which they are trained. This worry has intensified as the quality and complexity of generative models have immensely improved, and the availability of extensive datasets containing copyrighted material has expanded. Researchers are actively exploring strategies to mitigate the risk of generating infringing samples, with a recent line of work suggesting to employ techniques such as differential privacy and other forms of algorithmic stability to provide guarantees on the lack of infringing copying. In this work, we examine whether such algorithmic stability techniques are suitable to ensure the responsible use of generative models without inadvertently violating copyright laws. We argue that while these techniques aim to verify the presence of identifiable information in datasets, thus being privacy-oriented, copyright law aims to promote the use of original works for the benefit of society as a whole, provided that no unlicensed use of protected expression occurred. These fundamental differences between privacy and copyright must not be overlooked. In particular, we demonstrate that while algorithmic stability may be perceived as a practical tool to detect copying, such copying does not necessarily constitute copyright infringement. Therefore, if adopted as a standard for detecting an establishing copyright infringement, algorithmic stability may undermine the intended objectives of copyright law.
Successive interference cancellation (SIC) is used to approach the achievable information rates (AIRs) of joint detection and decoding for long-haul optical fiber links. The AIRs of memoryless ring constellations are compared to those of circularly symmetric complex Gaussian modulation for surrogate channel models with correlated phase noise. Simulations are performed for 1000 km of standard single-mode fiber with ideal Raman amplification. In this setup, 32 rings and 16 SIC-stages with Gaussian message-passing receivers achieve the AIR peaks of previous work. The computational complexity scales in proportion to the number of SIC-stages, where one stage has the complexity of separate detection and decoding.
Performance attribution analysis, defined as the process of explaining the drivers of the excess performance of an investment portfolio against a benchmark, stands as a significant feature of portfolio management and plays a crucial role in the investment decision-making process, particularly within the fund management industry. Rooted in a solid financial and mathematical framework, the importance and methodologies of this analytical technique are extensively documented across numerous academic research papers and books. The integration of large language models (LLMs) and AI agents marks a groundbreaking development in this field. These agents are designed to automate and enhance the performance attribution analysis by accurately calculating and analyzing portfolio performances against benchmarks. In this study, we introduce the application of an AI Agent for a variety of essential performance attribution tasks, including the analysis of performance drivers and utilizing LLMs as calculation engine for multi-level attribution analysis and question-answering (QA) tasks. Leveraging advanced prompt engineering techniques such as Chain-of-Thought (CoT) and Plan and Solve (PS), and employing a standard agent framework from LangChain, the research achieves promising results: it achieves accuracy rates exceeding 93% in analyzing performance drivers, attains 100% in multi-level attribution calculations, and surpasses 84% accuracy in QA exercises that simulate official examination standards. These findings affirm the impactful role of AI agents, prompt engineering and evaluation in advancing portfolio management processes, highlighting a significant development in the practical application and evaluation of Generative AI technologies within the domain.
The LSTM network was proposed to overcome the difficulty in learning long-term dependence, and has made significant advancements in applications. With its success and drawbacks in mind, this paper raises the question - do RNN and LSTM have long memory? We answer it partially by proving that RNN and LSTM do not have long memory from a statistical perspective. A new definition for long memory networks is further introduced, and it requires the model weights to decay at a polynomial rate. To verify our theory, we convert RNN and LSTM into long memory networks by making a minimal modification, and their superiority is illustrated in modeling long-term dependence of various datasets.
Compared with cheap addition operation, multiplication operation is of much higher computation complexity. The widely-used convolutions in deep neural networks are exactly cross-correlation to measure the similarity between input feature and convolution filters, which involves massive multiplications between float values. In this paper, we present adder networks (AdderNets) to trade these massive multiplications in deep neural networks, especially convolutional neural networks (CNNs), for much cheaper additions to reduce computation costs. In AdderNets, we take the $\ell_1$-norm distance between filters and input feature as the output response. The influence of this new similarity measure on the optimization of neural network have been thoroughly analyzed. To achieve a better performance, we develop a special back-propagation approach for AdderNets by investigating the full-precision gradient. We then propose an adaptive learning rate strategy to enhance the training procedure of AdderNets according to the magnitude of each neuron's gradient. As a result, the proposed AdderNets can achieve 74.9% Top-1 accuracy 91.7% Top-5 accuracy using ResNet-50 on the ImageNet dataset without any multiplication in convolution layer.
Language model pre-training has proven to be useful in learning universal language representations. As a state-of-the-art language model pre-training model, BERT (Bidirectional Encoder Representations from Transformers) has achieved amazing results in many language understanding tasks. In this paper, we conduct exhaustive experiments to investigate different fine-tuning methods of BERT on text classification task and provide a general solution for BERT fine-tuning. Finally, the proposed solution obtains new state-of-the-art results on eight widely-studied text classification datasets.
Recently, graph neural networks (GNNs) have revolutionized the field of graph representation learning through effectively learned node embeddings, and achieved state-of-the-art results in tasks such as node classification and link prediction. However, current GNN methods are inherently flat and do not learn hierarchical representations of graphs---a limitation that is especially problematic for the task of graph classification, where the goal is to predict the label associated with an entire graph. Here we propose DiffPool, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion. DiffPool learns a differentiable soft cluster assignment for nodes at each layer of a deep GNN, mapping nodes to a set of clusters, which then form the coarsened input for the next GNN layer. Our experimental results show that combining existing GNN methods with DiffPool yields an average improvement of 5-10% accuracy on graph classification benchmarks, compared to all existing pooling approaches, achieving a new state-of-the-art on four out of five benchmark data sets.
Deep neural networks (DNNs) have been found to be vulnerable to adversarial examples resulting from adding small-magnitude perturbations to inputs. Such adversarial examples can mislead DNNs to produce adversary-selected results. Different attack strategies have been proposed to generate adversarial examples, but how to produce them with high perceptual quality and more efficiently requires more research efforts. In this paper, we propose AdvGAN to generate adversarial examples with generative adversarial networks (GANs), which can learn and approximate the distribution of original instances. For AdvGAN, once the generator is trained, it can generate adversarial perturbations efficiently for any instance, so as to potentially accelerate adversarial training as defenses. We apply AdvGAN in both semi-whitebox and black-box attack settings. In semi-whitebox attacks, there is no need to access the original target model after the generator is trained, in contrast to traditional white-box attacks. In black-box attacks, we dynamically train a distilled model for the black-box model and optimize the generator accordingly. Adversarial examples generated by AdvGAN on different target models have high attack success rate under state-of-the-art defenses compared to other attacks. Our attack has placed the first with 92.76% accuracy on a public MNIST black-box attack challenge.
小貼士
登錄享
相關主題
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191