Search

The rapid growth in the parameters of large language models (LLMs) has made inference latency a fundamental bottleneck, limiting broader application of LLMs. Speculative decoding represents a lossless approach to accelerate inference through a guess-and-verify paradigm, leveraging the parallel capabilities of modern hardware. Some speculative decoding methods rely on additional structures to guess draft tokens, such as small models or parameter-efficient architectures, which need extra training before use. Alternatively, retrieval-based train-free techniques build libraries from pre-existing corpora or by n-gram generation. However, they face challenges like large storage requirements, time-consuming retrieval, and limited adaptability. Observing that candidate tokens generated during the decoding process are likely to reoccur in future sequences, we propose Token Recycling. This approach stores candidate tokens in an adjacency matrix and employs a breadth-first search (BFS)-like algorithm on the matrix to construct a draft tree. The tree is then validated through tree attention. New candidate tokens from the decoding process are then used to update the matrix. Token Recycling requires \textless2MB of additional storage and achieves approximately 2x speedup across all sizes of LLMs. It significantly outperforms existing train-free methods by 30\% and even a training method by 25\%. It can be directly applied to any existing LLMs and tasks without the need for adaptation., Comment: under review

This study introduces an innovative method for analyzing the impact of various interventions on customer churn, using the potential outcomes framework. We present a new causal model, the tensorized latent factor block hazard model, which incorporates tensor completion methods for a principled causal analysis of customer churn. A crucial element of our approach is the formulation of a 1-bit tensor completion for the parameter tensor. This captures hidden customer characteristics and temporal elements from churn records, effectively addressing the binary nature of churn data and its time-monotonic trends. Our model also uniquely categorizes interventions by their similar impacts, enhancing the precision and practicality of implementing customer retention strategies. For computational efficiency, we apply a projected gradient descent algorithm combined with spectral clustering. We lay down the theoretical groundwork for our model, including its non-asymptotic properties. The efficacy and superiority of our model are further validated through comprehensive experiments on both simulated and real-world applications., Comment: Accepted for publication in ICML, 2024

Semi-quantum key distribution (SQKD) allows sharing random keys between a quantum user and a classical user. However, implementing classical user operations is challenging, posing a hurdle to achieving the Single-state protocol. By using the "selective modulation" method, the feasibility of SQKD is verified in principle. The proposal of the selective modulation method enables the realization of other protocols for SQKD. To advance experimental progress in SQKD, we propose and implement a phase-encoded semi-quantum key distribution system based on the Single-state protocol and the "selective modulation" method. The system operates at a frequency of 100MHz and an average photon number of 0.1. The interference contrast achieved 96.52%, the average quantum bit error rate was 1.19%, and the raw key rate reached 88Kbps. Our experimental results demonstrate the feasibility and stability of the proposed phase-encoded semi-quantum key distribution system. Furthermore, by leveraging the "selective modulation" scheme proposed in this paper, we develop a comprehensive theoretical description of selective modulation. Through an analysis of quantum state evolution, we assess the security of our system, ultimately demonstrating its resilience against attacks targeting quantum states. The classical user of our system requires only two optical devices, significantly reducing the equipment requirements and enhancing its application potential. This work validates the feasibility of semi-quantum key distribution experiments and provides ideas for future research on semi-quantum key distribution experiments and security studies.

Instruction tuning plays a pivotal role in Code Large Language Models (Code LLMs) for the task of program synthesis. Presently, two dominant paradigms for collecting tuning data are natural-instruct (human-written) and self-instruct (automatically generated). Natural-instruct includes diverse and correct codes but lacks instruction-code pairs, and exists improper code formats like nested single-line codes. In contrast, self-instruct automatically generates proper paired data. However, it suffers from low diversity due to generating duplicates and cannot ensure the correctness of codes. To bridge the both paradigms, we propose \textbf{Semi-Instruct}. It first converts diverse but improper codes from natural-instruct into proper instruction-code pairs through a method similar to self-instruct. To verify the correctness of generated codes, we design a novel way to construct test cases by generating cases' inputs and executing correct codes from natural-instruct to get outputs. Finally, diverse and correct instruction-code pairs are retained for instruction tuning. Experiments show that semi-instruct is significantly better than natural-instruct and self-instruct. Furthermore, the performance steadily improves as data scale increases.

Program of Thoughts (PoT) is an approach characterized by its executable intermediate steps, which ensure the accuracy of the logical calculations in the reasoning process. Currently, PoT primarily uses Python. However, relying solely on a single language may result in suboptimal solutions and overlook the potential benefits of other programming languages. In this paper, we conduct comprehensive experiments on the programming languages used in PoT and find that no single language consistently delivers optimal performance across all tasks and models. The effectiveness of each language varies depending on the specific scenarios. Inspired by this, we propose a task and model agnostic approach called MultiPoT, which harnesses strength and diversity from various languages. Experimental results reveal that it significantly outperforms Python Self-Consistency. Furthermore, it achieves comparable or superior performance compared to the best monolingual PoT in almost all tasks across all models. In particular, MultiPoT achieves more than 4.6% improvement on average on ChatGPT (gpt-3.5-turbo-0701)., Comment: under review

In recent years, real image super-resolution (SR) has achieved promising results due to the development of SR datasets and corresponding real SR methods. In contrast, the field of real video SR is lagging behind, especially for real raw videos. Considering the superiority of raw image SR over sRGB image SR, we construct a real-world raw video SR (Real-RawVSR) dataset and propose a corresponding SR method. We utilize two DSLR cameras and a beam-splitter to simultaneously capture low-resolution (LR) and high-resolution (HR) raw videos with 2x, 3x, and 4x magnifications. There are 450 video pairs in our dataset, with scenes varying from indoor to outdoor, and motions including camera and object movements. To our knowledge, this is the first real-world raw VSR dataset. Since the raw video is characterized by the Bayer pattern, we propose a two-branch network, which deals with both the packed RGGB sequence and the original Bayer pattern sequence, and the two branches are complementary to each other. After going through the proposed co-alignment, interaction, fusion, and reconstruction modules, we generate the corresponding HR sRGB sequence. Experimental results demonstrate that the proposed method outperforms benchmark real and synthetic video SR methods with either raw or sRGB inputs. Our code and dataset are available at https://github.com/zmzhang1998/Real-RawVSR., Comment: Accepted by ECCV2022

Machine learning approaches have been widely used for discovering the underlying physics of dynamical systems from measured data. Existing approaches, however, still lack robustness, especially when the measured data contain a large level of noise. The lack of robustness is mainly attributed to the insufficient representativeness of used features. As a result, the intrinsic mechanism governing the observed system cannot be accurately identified. In this study, we use an efficient topological descriptor for complex data, i.e., the Euler characteristics (ECs), as features to characterize the spatiotemporal data collected from dynamical systems and discover the underlying physics. Unsupervised manifold learning and supervised classification results show that EC can be used to efficiently distinguish systems with different while similar governing models. We also demonstrate that the machine learning approaches using EC can improve the confidence level of sparse regression methods of physics discovery.

Robust physics discovery is of great interest for many scientific and engineering fields. Inspired by the principle that a representative model is the one simplest possible, a new model selection criteria considering both model's Parsimony and Sparsity is proposed. A Parsimony Enhanced Sparse Bayesian Learning (PeSBL) method is developed for discovering the governing Partial Differential Equations (PDEs) of nonlinear dynamical systems. Compared with the conventional Sparse Bayesian Learning (SBL) method, the PeSBL method promotes parsimony of the learned model in addition to its sparsity. In this method, the parsimony of model terms is evaluated using their locations in the prescribed candidate library, for the first time, considering the increased complexity with the power of polynomials and the order of spatial derivatives. Subsequently, the model parameters are updated through Bayesian inference with the raw data. This procedure aims to reduce the error associated with the possible loss of information in data preprocessing and numerical differentiation prior to sparse regression. Results of numerical case studies indicate that the governing PDEs of many canonical dynamical systems can be correctly identified using the proposed PeSBL method from highly noisy data (up to 50% in the current study). Next, the proposed methodology is extended for stochastic PDE learning where all parameters and modeling error are considered as random variables. Hierarchical Bayesian Inference (HBI) is integrated with the proposed framework for stochastic PDE learning from a population of observations. Finally, the proposed PeSBL is demonstrated for system response prediction with uncertainties and anomaly diagnosis. Codes of all demonstrated examples in this study are available on the website: https://github.com/ymlasu., Comment: arXiv admin note: text overlap with arXiv:2102.06504