arXiv subjects

Xiaoyu Chen

Publications and source records attributed to Xiaoyu Chen.

At least 109 records · Page 6Linked to original sources

arXiv2024

Asking Before Acting: Gather Information in Embodied Decision Making with Language Models

With strong capabilities of reasoning and a broad understanding of the world, Large Language Models (LLMs) have demonstrated immense potential in building versatile embodied decision-making agents capable of executing a wide array of tasks. Nevertheless, when deployed in unfamiliar environments, we show that LLM agents encounter challenges in efficiently gathering essential information, leading to suboptimal performance. Conversely, human individuals often seek additional information from their peers prior to taking action, harnessing external knowledge to avoid unnecessary trial and error. Drawing inspiration from this behavior, we propose \textit{Asking Before Acting} (ABA), a method that empowers the agent to proactively inquire with external sources for pertinent information using natural language during their interactions within the environment. In this way, the agent is able to enhance its efficiency and performance by circumventing potentially laborious steps and combating the difficulties associated with exploration in unfamiliar environments and vagueness of the instructions. We conduct extensive experiments involving a spectrum of environments including text-based household everyday tasks, robot arm manipulation tasks, and real world open domain image based embodied tasks. The experiments involve various models from Vicuna to GPT-4. The results demonstrate that, even with modest prompts modifications, ABA exhibits substantial advantages on both performance and efficiency over baseline LLM agents. Further finetuning ABA with reformulated metadata (ABA-FT) faciliates learning the rationale for asking and allows for additional enhancements especially in tasks that baselines struggle to solve.

cs.AI↗

arXiv2024

A Novel Paradigm Boosting Translation Capabilities of Large Language Models

This paper presents a study on strategies to enhance the translation capabilities of large language models (LLMs) in the context of machine translation (MT) tasks. The paper proposes a novel paradigm consisting of three stages: Secondary Pre-training using Extensive Monolingual Data, Continual Pre-training with Interlinear Text Format Documents, and Leveraging Source-Language Consistent Instruction for Supervised Fine-Tuning. Previous research on LLMs focused on various strategies for supervised fine-tuning (SFT), but their effectiveness has been limited. While traditional machine translation approaches rely on vast amounts of parallel bilingual data, our paradigm highlights the importance of using smaller sets of high-quality bilingual data. We argue that the focus should be on augmenting LLMs' cross-lingual alignment abilities during pre-training rather than solely relying on extensive bilingual data during SFT. Experimental results conducted using the Llama2 model, particularly on Chinese-Llama2 after monolingual augmentation, demonstrate the improved translation capabilities of LLMs. A significant contribution of our approach lies in Stage2: Continual Pre-training with Interlinear Text Format Documents, which requires less than 1B training data, making our method highly efficient. Additionally, in Stage3, we observed that setting instructions consistent with the source language benefits the supervised fine-tuning process. Experimental results demonstrate that our approach surpasses previous work and achieves superior performance compared to models such as NLLB-54B and GPT3.5-text-davinci-003, despite having a significantly smaller parameter count of only 7B or 13B. This achievement establishes our method as a pioneering strategy in the field of machine translation.

cs.CL↗

arXiv2024

Spectral Independence Beyond Total Influence on Trees and Related Graphs

We study how to establish $\textit{spectral independence}$, a key concept in sampling, without relying on total influence bounds, by applying an $\textit{approximate inverse}$ of the influence matrix. Our method gives constant upper bounds on spectral independence for two foundational Gibbs distributions known to have unbounded total influences: $\bullet$ The monomer-dimer model on graphs with large girth (including trees). Prior to our work, such results were only known for graphs with constant maximum degrees or infinite regular trees, as shown by Chen, Liu, and Vigoda (STOC '21). $\bullet$ The hardcore model on trees with fugacity $λ< \mathrm{e}^2$. This remarkably surpasses the well-known $λ_r>\mathrm{e}-1$ lower bound for the reconstruction threshold on trees, significantly improving upon the current threshold $λ< 1.3$, established in a prior work by Efthymiou, Hayes, Štefankovič, and Vigoda (RANDOM '23). Consequently, we establish optimal $Ω(n^{-1})$ spectral gaps of the Glauber dynamics for these models on arbitrary trees, regardless of the maximum degree $Δ$.

cs.DS↗

arXiv2024

An edge detection-based deep learning approach for tear meniscus height measurement

Automatic measurements of tear meniscus height (TMH) have been achieved by using deep learning techniques; however, annotation is significantly influenced by subjective factors and is both time-consuming and labor-intensive. In this paper, we introduce an automatic TMH measurement technique based on edge detection-assisted annotation within a deep learning framework. This method generates mask labels less affected by subjective factors with enhanced efficiency compared to previous annotation approaches. For improved segmentation of the pupil and tear meniscus areas, the convolutional neural network Inceptionv3 was first implemented as an image quality assessment model, effectively identifying higher-quality images with an accuracy of 98.224%. Subsequently, by using the generated labels, various algorithms, including Unet, ResUnet, Deeplabv3+FcnResnet101, Deeplabv3+FcnResnet50, FcnResnet50, and FcnResnet101 were trained, with Unet demonstrating the best performance. Finally, Unet was used for automatic pupil and tear meniscus segmentation to locate the center of the pupil and calculate TMH,respectively. An evaluation of the mask quality predicted by Unet indicated a Mean Intersection over Union of 0.9362, a recall of 0.9261, a precision of 0.9423, and an F1-Score of 0.9326. Additionally, the TMH predicted by the model was assessed, with the fitting curve represented as y= 0.982x-0.862, an overall correlation coefficient of r^2=0.961 , and an accuracy of 94.80% (237/250). In summary, the algorithm can automatically screen images based on their quality,segment the pupil and tear meniscus areas, and automatically measure TMH. Measurement results using the AI algorithm demonstrate a high level of consistency with manual measurements, offering significant support to clinical doctors in diagnosing dry eye disease.

eess.IV↗

arXiv2024

Identifying possible mechanism for quantum needle in chemical magnetoreception

The radical pair mechanism is an important model that may provide a basis for biological magnetoreception. To account for the high orientation precision of the real avian compass, P. J. Hore et al. proposed an intriguing phenomenon called quantum needle [Proc. Natl. Acad. Sci. 113, 4634 (2016)], where a spike-like feature emerges in the fractional yield signal. However, it is believed that quantum needle requires the radical pair lifetime to be longer than a few microseconds and thus poses stern challenges in realistic biological systems. Here, we exploit the optimization techniques and find a novel class of model system, which sustains much more prominent features of quantum needle and significantly relaxes the requirement for radical pair lifetime. Even more surprisingly, we find that the characteristics of quantum needle retain a narrow functional window around the geomagnetic field, which is absent in the previous model systems. Therefore, our work provides essential evidence for identifying the possible physical mechanism for quantum needle in chemical magnetoreception.

physics.bio-ph↗

arXiv2024

R-BI: Regularized Batched Inputs enhance Incremental Decoding Framework for Low-Latency Simultaneous Speech Translation

Incremental Decoding is an effective framework that enables the use of an offline model in a simultaneous setting without modifying the original model, making it suitable for Low-Latency Simultaneous Speech Translation. However, this framework may introduce errors when the system outputs from incomplete input. To reduce these output errors, several strategies such as Hold-$n$, LA-$n$, and SP-$n$ can be employed, but the hyper-parameter $n$ needs to be carefully selected for optimal performance. Moreover, these strategies are more suitable for end-to-end systems than cascade systems. In our paper, we propose a new adaptable and efficient policy named "Regularized Batched Inputs". Our method stands out by enhancing input diversity to mitigate output errors. We suggest particular regularization techniques for both end-to-end and cascade systems. We conducted experiments on IWSLT Simultaneous Speech Translation (SimulST) tasks, which demonstrate that our approach achieves low latency while maintaining no more than 2 BLEU points loss compared to offline systems. Furthermore, our SimulST systems attained several new state-of-the-art results in various language directions.

cs.CL↗

arXiv2023

INarIG: Iterative Non-autoregressive Instruct Generation Model For Word-Level Auto Completion

Computer-aided translation (CAT) aims to enhance human translation efficiency and is still important in scenarios where machine translation cannot meet quality requirements. One fundamental task within this field is Word-Level Auto Completion (WLAC). WLAC predicts a target word given a source sentence, translation context, and a human typed character sequence. Previous works either employ word classification models to exploit contextual information from both sides of the target word or directly disregarded the dependencies from the right-side context. Furthermore, the key information, i.e. human typed sequences, is only used as prefix constraints in the decoding module. In this paper, we propose the INarIG (Iterative Non-autoregressive Instruct Generation) model, which constructs the human typed sequence into Instruction Unit and employs iterative decoding with subwords to fully utilize input information given in the task. Our model is more competent in dealing with low-frequency words (core scenario of this task), and achieves state-of-the-art results on the WMT22 and benchmark datasets, with a maximum increase of over 10% prediction accuracy.

cs.CL↗

arXiv2023

Ensemble Active Learning by Contextual Bandits for AI Incubation in Manufacturing

It is challenging but important to save annotation efforts in streaming data acquisition to maintain data quality for supervised learning base learners. We propose an ensemble active learning method to actively acquire samples for annotation by contextual bandits, which is will enforce the exploration-exploitation balance and leading to improved AI modeling performance.

cs.LG↗

arXiv2023

Near-linear time samplers for matroid independent sets with applications

We give a $\widetilde{O}(n)$ time almost uniform sampler for independent sets of a matroid, whose ground set has $n$ elements and is given by an independence oracle. As a consequence, one can sample connected spanning subgraphs of a given graph $G=(V,E)$ in $\widetilde{O}(|E|)$ time. This leads to improved running time on estimating all-terminal network reliability. Furthermore, we generalise this near-linear time sampler to the random cluster model with $q\le 1$.

cs.DS↗

arXiv2023

A Near-Linear Time Sampler for the Ising Model with External Field

We give a near-linear time sampler for the Gibbs distribution of the ferromagnetic Ising models with edge activities $\boldsymbolβ > 1$ and external fields $\boldsymbolλ<1$ (or symmetrically, $\boldsymbolλ>1$) on general graphs with bounded or unbounded maximum degree. Our algorithm is based on the field dynamics given in [CFYZ21]. We prove the correctness and efficiency of our algorithm by establishing spectral independence of distribution of the random cluster model and the rapid mixing of Glauber dynamics on the random cluster model in a low-temperature regime, which may be of independent interest.

math.PR↗

arXiv2023

Auditory Attention Decoding with Task-Related Multi-View Contrastive Learning

The human brain can easily focus on one speaker and suppress others in scenarios such as a cocktail party. Recently, researchers found that auditory attention can be decoded from the electroencephalogram (EEG) data. However, most existing deep learning methods are difficult to use prior knowledge of different views (that is attended speech and EEG are task-related views) and extract an unsatisfactory representation. Inspired by Broadbent's filter model, we decode auditory attention in a multi-view paradigm and extract the most relevant and important information utilizing the missing view. Specifically, we propose an auditory attention decoding (AAD) method based on multi-view VAE with task-related multi-view contrastive (TMC) learning. Employing TMC learning in multi-view VAE can utilize the missing view to accumulate prior knowledge of different views into the fusion of representation, and extract the approximate task-related representation. We examine our method on two popular AAD datasets, and demonstrate the superiority of our method by comparing it to the state-of-the-art method.

cs.SD↗

arXiv2023

On the Power of Pre-training for Generalization in RL: Provable Benefits and Hardness

Generalization in Reinforcement Learning (RL) aims to learn an agent during training that generalizes to the target environment. This paper studies RL generalization from a theoretical aspect: how much can we expect pre-training over training environments to be helpful? When the interaction with the target environment is not allowed, we certify that the best we can obtain is a near-optimal policy in an average sense, and we design an algorithm that achieves this goal. Furthermore, when the agent is allowed to interact with the target environment, we give a surprising result showing that asymptotically, the improvement from pre-training is at most a constant factor. On the other hand, in the non-asymptotic regime, we design an efficient algorithm and prove a distribution-based regret bound in the target environment that is independent of the state-action space.

cs.LG↗

arXiv2023

Near-Optimal Quantum Coreset Construction Algorithms for Clustering

$k$-Clustering in $\mathbb{R}^d$ (e.g., $k$-median and $k$-means) is a fundamental machine learning problem. While near-linear time approximation algorithms were known in the classical setting for a dataset with cardinality $n$, it remains open to find sublinear-time quantum algorithms. We give quantum algorithms that find coresets for $k$-clustering in $\mathbb{R}^d$ with $\tilde{O}(\sqrt{nk}d^{3/2})$ query complexity. Our coreset reduces the input size from $n$ to $\mathrm{poly}(kε^{-1}d)$, so that existing $α$-approximation algorithms for clustering can run on top of it and yield $(1 + ε)α$-approximation. This eventually yields a quadratic speedup for various $k$-clustering approximation algorithms. We complement our algorithm with a nearly matching lower bound, that any quantum algorithm must make $Ω(\sqrt{nk})$ queries in order to achieve even $O(1)$-approximation for $k$-clustering.

quant-ph↗

arXiv2023

Text Style Transfer Back-Translation

Back Translation (BT) is widely used in the field of machine translation, as it has been proved effective for enhancing translation quality. However, BT mainly improves the translation of inputs that share a similar style (to be more specific, translation-like inputs), since the source side of BT data is machine-translated. For natural inputs, BT brings only slight improvements and sometimes even adverse effects. To address this issue, we propose Text Style Transfer Back Translation (TST BT), which uses a style transfer model to modify the source side of BT data. By making the style of source-side text more natural, we aim to improve the translation of natural inputs. Our experiments on various language pairs, including both high-resource and low-resource ones, demonstrate that TST BT significantly improves translation performance against popular BT benchmarks. In addition, TST BT is proved to be effective in domain adaptation so this strategy can be regarded as a general data augmentation method. Our training code and text style transfer model are open-sourced.

cs.CL↗

arXiv2023

Towards Generalizable Reinforcement Learning for Trade Execution

Optimized trade execution is to sell (or buy) a given amount of assets in a given time with the lowest possible trading cost. Recently, reinforcement learning (RL) has been applied to optimized trade execution to learn smarter policies from market data. However, we find that many existing RL methods exhibit considerable overfitting which prevents them from real deployment. In this paper, we provide an extensive study on the overfitting problem in optimized trade execution. First, we model the optimized trade execution as offline RL with dynamic context (ORDC), where the context represents market variables that cannot be influenced by the trading policy and are collected in an offline manner. Under this framework, we derive the generalization bound and find that the overfitting issue is caused by large context space and limited context samples in the offline setting. Accordingly, we propose to learn compact representations for context to address the overfitting problem, either by leveraging prior knowledge or in an end-to-end manner. To evaluate our algorithms, we also implement a carefully designed simulator based on historical limit order book (LOB) data to provide a high-fidelity benchmark for different algorithms. Our experiments on the high-fidelity simulator demonstrate that our algorithms can effectively alleviate overfitting and achieve better performance.

q-fin.TR↗

arXiv2023

Uniqueness and Rapid Mixing in the Bipartite Hardcore Model

We characterize the uniqueness condition in the hardcore model for bipartite graphs with degree bounds only on one side, and provide a nearly linear time sampling algorithm that works up to the uniqueness threshold. We show that the uniqueness threshold for bipartite graph has almost the same form of the tree uniqueness threshold for general graphs, except with degree bounds only on one side of the bipartition. The hardcore model from statistical physics can be seen as a weighted enumeration of independent sets. Its bipartite version (#BIS) is a central open problem in approximate counting. Compared to the same problem in a general graph, surprising tractable regime have been identified that are believed to be hard in general. This is made possible by two lines of algorithmic approach: the high-temperature algorithms starting from Liu and Lu (STOC 2015), and the low-temperature algorithms starting from Helmuth, Perkins, and Regts (STOC 2019). In this work, we study the limit of these algorithms in the high-temperature case. Our characterization of the uniqueness condition is obtained by proving decay of correlations for arguably the best possible regime, which involves locating fixpoints of multivariate iterative rational maps and showing their contraction. We also give a nearly linear time sampling algorithm based on simulating field dynamics only on one side of the bipartite graph that works up to the uniqueness threshold. Our algorithm is very different from the original high-temperature algorithm of Liu and Lu, and it makes use of a connection between correlation decay and spectral independence of Markov chains. Last but not the least, we are able to show that the standard Glauber dynamics on both side of the bipartite graph mixes in polynomial time up to the uniqueness.

cs.DS↗

arXiv2023

Streaming Euclidean Max-Cut: Dimension vs Data Reduction

Max-Cut is a fundamental problem that has been studied extensively in various settings. We design an algorithm for Euclidean Max-Cut, where the input is a set of points in $\mathbb{R}^d$, in the model of dynamic geometric streams, where the input $X\subseteq [Δ]^d$ is presented as a sequence of point insertions and deletions. Previously, Frahling and Sohler [STOC 2005] designed a $(1+ε)$-approximation algorithm for the low-dimensional regime, i.e., it uses space $\exp(d)$. To tackle this problem in the high-dimensional regime, which is of growing interest, one must improve the dependence on the dimension $d$, ideally to space complexity $\mathrm{poly}(ε^{-1} d \logΔ)$. Lammersen, Sidiropoulos, and Sohler [WADS 2009] proved that Euclidean Max-Cut admits dimension reduction with target dimension $d' = \mathrm{poly}(ε^{-1})$. Combining this with the aforementioned algorithm that uses space $\exp(d')$, they obtain an algorithm whose overall space complexity is indeed polynomial in $d$, but unfortunately exponential in $ε^{-1}$. We devise an alternative approach of \emph{data reduction}, based on importance sampling, and achieve space bound $\mathrm{poly}(ε^{-1} d \logΔ)$, which is exponentially better (in $ε$) than the dimension-reduction approach. To implement this scheme in the streaming model, we employ a randomly-shifted quadtree to construct a tree embedding. While this is a well-known method, a key feature of our algorithm is that the embedding's distortion $O(d\logΔ)$ affects only the space complexity, and the approximation ratio remains $1+ε$.

cs.DS↗

arXiv2023

TBFormer: Two-Branch Transformer for Image Forgery Localization

Image forgery localization aims to identify forged regions by capturing subtle traces from high-quality discriminative features. In this paper, we propose a Transformer-style network with two feature extraction branches for image forgery localization, and it is named as Two-Branch Transformer (TBFormer). Firstly, two feature extraction branches are elaborately designed, taking advantage of the discriminative stacked Transformer layers, for both RGB and noise domain features. Secondly, an Attention-aware Hierarchical-feature Fusion Module (AHFM) is proposed to effectively fuse hierarchical features from two different domains. Although the two feature extraction branches have the same architecture, their features have significant differences since they are extracted from different domains. We adopt position attention to embed them into a unified feature domain for hierarchical feature investigation. Finally, a Transformer decoder is constructed for feature reconstruction to generate the predicted mask. Extensive experiments on publicly available datasets demonstrate the effectiveness of the proposed model.

cs.CV↗