检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Ali-Dib, Mohamad Menou, Kristen New York University Abu Dhabi PO Box 129188 United Arab Emirates Physics & Astrophysics Group DPES University of Toronto ScarboroughONM1C 1A4 Canada Department of Astronomy & Astrophysics University of Toronto ONM5S 3H4 Canada Department of Physics University of Toronto ONM5S 1A7 Canada Observatoire de la Cote d’Azur TOP team Laboratoire Lagrange - CNRS Nice France

Large Language Models (LLMs) can solve some undergraduate-level to graduate-level physics textbook problems and are proficient at coding. Combining these two capabilities could one day enable AI systems to simulate and predict the physical world. We present an evaluation of state-of-the-art (SOTA) LLMs on PhD-level to research-level computational physics problems. We condition LLM generation on the use of well-documented and widely-used packages to elicit coding capabilities in the physics and astrophysics domains. We contribute ∼ 50 original and challenging problems in celestial mechanics (with REBOUND), stellar physics (with MESA), 1D fluid dynamics (with Dedalus) and non-linear dynamics (with SciPy). Since our problems do not admit unique solutions, we evaluate LLM performance on several soft metrics: counts of lines that contain different types of errors (coding, physics, necessity and sufficiency) as well as a more ‘educational’ Pass-Fail metric focused on capturing the salient physical ingredients of the problem at hand. As expected, today’s SOTA LLM (GPT4) zero-shot fails most of our problems, although about 40% of the solutions could plausibly get a passing grade. About 70 − 90% of the code lines produced are necessary, sufficient and correct (coding & physics). Physics and coding errors are the most common, with some unnecessary or insufficient lines. We observe significant variations across problem class and difficulty. We identify several failure modes of GPT4 in the computational physics domain, such as poor physical units handling, poor code versioning, tendency to hallucinate plausible sub-modules, lack of physical justification for global run parameters (e.g., simulation time, or upper-lower bounds for parametric exploration) and inability to define steady-state or stopping conditions reliably. Our reconnaissance work provides a snapshot of current computational capabilities in classical physics and points to obvious improvement targets if AI systems

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

Asymptotically Optimal Codes for (t,s)-Burst Error

arXiv

引用

arXiv 2024年

作者： Sun, Yubo Lu, Ziyang Zhang, Yiwei Ge, Gennian School of Mathematical Sciences Capital Normal University Beijing100048 China Key Laboratory of Cryptologic Technology and Information Security Ministry of Education China School of Cyber Science and Technology Shandong University Shandong Qingdao266237 China

Recently, codes for correcting a burst of errors have attracted significant attention. One of the most important reasons is that bursts of errors occur in certain emerging techniques, such as DNA storage. In this paper, we investigate a type of error, called a (t, s)-burst, which deletes t consecutive symbols and inserts s arbitrary symbols at the same coordinate. Note that a (t, s)-burst error can be seen as a generalization of a burst of insertions (t = 0), a burst of deletions (s = 0), and a burst of substitutions (t = s). Our main contribution is to give explicit constructions of q-ary (t, s)-burst correcting codes with log n + O(1) bits of redundancy for any given constant non-negative integers t, s, and q ≥ 2. These codes have optimal redundancy up to an additive constant. Furthermore, we apply our (t, s)-burst correcting codes to combat other various types of errors and improve the corresponding results. In particular, one of our byproducts is a permutation code capable of correcting a burst of t stable deletions with log n + O(1) bits of redundancy, which is optimal up to an additive constant. Copyright © 2024, The Authors. All rights reserved.

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

MULTI-PROGRAMMING LANGUAGE ENSEMBLE FOR CODE GENERATION IN LARGE LANGUAGE MODEL

arXiv

引用

arXiv 2024年

作者： Xue, Tengfei Li, Xuefeng Azim, Tahir Smirnov, Roman Yu, Jianhui Sadrieh, Arash Pahlavan, Babak NinjaTech AI United States

Large language models (LLMs) have significantly improved code generation, particularly in one-pass code generation. However, most existing approaches focus solely on generating code in a single programming language, overlooking the potential of leveraging the multi-language capabilities of LLMs. LLMs have varying patterns of errors across different languages, suggesting that a more robust approach could be developed by leveraging these multi-language outputs. In this study, we propose Multi-Programming Language Ensemble (MPLE), a novel ensemble-based method that utilizes code generation across multiple programming languages to enhance overall performance. By treating each language-specific code generation process as an individual "weak expert" and effectively integrating their outputs, our method mitigates language-specific errors and biases. This multi-language ensemble strategy leverages the complementary strengths of different programming languages, enabling the model to produce more accurate and robust code. Our approach can be seamlessly integrated with commonly used techniques such as the reflection algorithm and Monte Carlo tree search to improve code generation quality further. Experimental results show that our framework consistently enhances baseline performance by up to 17.92% on existing benchmarks (HumanEval and HumanEval-plus), with a standout result of 96.25% accuracy on the HumanEval benchmark, achieving new state-of-the-art results across various LLM models. The code will be released at https://***/NinjaTechAI/MPLE © 2024, CC BY.

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

Content coding and Search for Risk Assessment: Problems and Solutions

Content Coding and Search for Risk Assessment: Problems and ...

引用

IEEE Aerospace Conference

作者： Jane T. Malin Carroll Thronesbery David R. Throop 1610 Wavecrest Lane Houston TX 77062 S&K Global Solutions 1331 Gemini Suite 315 Houston TX 77058 The Boeing Company 2100 Space Park Drive Houston TX 77058

ISBN: (纸本)9781467376778

Risk assessment depends on data from problem reports (PRs), yet analysis of these reports is difficult. Finding all the pertinent reports is a challenge. Full-text search is commonly used to retrieve groups of similar PRs, but this method typically misses many reports because of variations in spelling, synonyms and technical jargon. Metadata codes are used to classify the content of each PR according to features such as the type of failure mode, defect or cause. However, manual PR coding schemes are difficult to follow and there are frequent coding errors. NASA reports have made recommendations to improve the quality of content codes, but these recommendations have not been adequately addressed. During an assessment for the International Space Station (ISS), alternative solutions to metadata coding were explored. Automatic concept extraction from the content of text descriptions was used to assign concept tags to each PR The concepts were drawn from large ontology of aerospace language. An automatic rule-based approach was developed to generate interoperable "proxy" codes. The rules associated the extracted concepts with manual PR codes. The work to automatically define proxy codes was abandoned because of the poor quality of the manual content codes. Lessons learned and recommended corrective and preventive actions for manual coding errors are discussed. They include improvements in development, definition, evaluation, review and correction of content coding errors and coding schemes. Automated concept extraction is a viable solution to the poor quality of manual PR codes. The benefits of this approach are discussed.

关键词： benefits approach discussed Codes Public Relations Praseodymium coding errors Solutions Risk Assessment Risk Management Metadata manual Interoperability AUTOMATIC

来源：评论

学校读者我要写书评

暂无评论

CODELUTRA: Boosting LLM Code Generation via Preference-Guided Refinement

arXiv

引用

arXiv 2024年

作者： Tao, Leitian Chen, Xiang Yu, Tong Mai, Tung Rossi, Ryan Li, Yixuan Mitra, Saayan University of Wisconsin-Madison United States Adobe Research United States

Large Language Models (LLMs) have revolutionized code generation but are require significant resources and tend to over-generalize, limiting their task-specific efficiency. Fine-tuning smaller, open-source LLMs is a cost-effective alternative, yet standard supervised approaches rely solely on correct examples, overlooking valuable insights from failures. We introduce CODELUTRA, a new framework that leverages both correct and incorrect code attempts. Instead of purely instructing with correct solutions, CODELUTRA uses iterative preference-based refinement, comparing successful and failed outputs to better approximate desired results. This process narrows the performance gap with state-of-the-art, larger models, without requiring massive datasets or auxiliary models. For example, on a challenging data science coding task, using only 500 samples improved Llama-3-8B’s accuracy from 28.2% to 48.6%, approaching GPT-4’s level. By capitalizing on both successes and mistakes, CODELUTRA offers a scalable, efficient path to high-quality code generation, making smaller open-source models more competitive with leading closed-source alternatives. © 2024, CC BY.

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

CODES ON WEIGHTED PROJECTIVE PLANES

arXiv

引用

arXiv 2024年

作者： Çakiroglu, Yagmur Nardi, Jade Şahin, Mesut Department of Mathematics Hacettepe University Ankara Turkey CNRS IRMAR Univ Rennes France

We comprehensively study weighted projective Reed-Muller (WPRM) codes on weighted projective planes P(1, a, b). We provide the universal Gröbner basis for the vanishing ideal of the set Y of Fq–rational points of P(1, a, b) to get the dimension of the code. We determine the regularity set of Y using a novel combinatorial approach. We employ footprint techniques to compute the minimum distance. Copyright © 2024, The Authors. All rights reserved.

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

On the Error-correcting Capability of Twisted Centralizer Codes Obtained from a Fixed Rank-1 Matrix

arXiv

引用

arXiv 2024年

作者： Temones, John Ben S. Central Bicol State University of Agriculture Philippines

Twisted centralizer codes are linear codes obtained by fixing a square matrix A and collecting all square matrices B satisfying the equality AB=aBA, where a is a fixed arbitrary constant. We denote the collection of such B’s as C(A,a), read as the centralizer of A, twisted by a;with every individual B’s as its codewords. In this paper, we give a generalization on the error-correcting capability of twisted centralizer codes obtained from a fixed rank-1 matrix. In particular, we fix the combinatorial matrix which is obtained by getting the linear combination of square 1’s matrix and identity matrix of order n. Results reveal that such codes have a dimension 1 for any fixed combinatorial matrix and constant a, hence having a relatively low information rate due to the way its codewords are constructed. However, the same family of codes are found to be maximum distance separable (MDS) codes, and hence can detect up to n2 - 1 errors and can correct up to floor(n2-12) errors, both the largest *** Codes coding theory © 2024, CC BY.

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

Mitigating Temporal Fragility in the XY Surface Code

arXiv

引用

arXiv 2023年

作者： Tsai, Pei-Kai Wu, Yue Puri, Shruti Yale University Department of Applied Physics New HavenCT06511 United States Yale Quantum Institute Yale University New HavenCT06511 United States Yale University Department of Computer Science New HavenCT06511 United States

An important outstanding challenge that must be overcome in order to fully utilize the XY surface code for correcting biased Pauli noise is the phenomenon of fragile temporal boundaries that arises during the standard logical state preparation and measurement protocols. To address this challenge we propose a new logical state preparation protocol based on locally entangling qubits into small Greenberger–Horne–Zeilinger-like states prior to making the stabilizer measurements that place them in the XY-code state. We prove that in this new procedure O(√n) high-rate errors along a single lattice boundary can cause a logical failure, leading to an almost quadratic reduction in the number of fault-configurations compared to the standard state-preparation approach. Moreover, the code becomes equivalent to a repetition code for high-rate errors, guaranteeing a 50% code-capacity threshold during state preparation for infinitely biased noise. With a simple matching decoder we confirm that our preparation protocol outperforms the standard one in terms of both threshold and logical error rate in the fault-tolerant regime where measurements are unreliable and at experimentally realistic biases. We also discuss how our state-preparation protocol can be inverted for similar fragile-boundary-mitigated logical-state measurement. © 2023, CC BY.

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

Identification over Permutation Channels

arXiv

引用

arXiv 2024年

作者： Sarkar, Abhishek Dey, Bikash Kumar Department of Electrical Engineering Indian Institute of Technology Bombay India

We study message identification over a q-ary uniform permutation channel, where the transmitted vector is permuted by a permutation chosen uniformly at random. For discrete memoryless channels (DMCs), the number of identifiable messages grows doubly exponentially. Identification capacity, the maximum second-order exponent, is known to be the same as the Shannon capacity of the DMC. Permutation channels support reliable communication of only polynomially many messages. A simple achievability result shows that message sizes growing as 2ǫnnq−1 are identifiable for any ǫn → 0. We prove two converse results. A "soft" converse shows that for any R > 0, there is no sequence of identification codes with message size growing as 2Rnq−1 with a power-law decay (n−µ) of the error probability. We also prove a "strong" converse showing that for any sequence of identification codes with message size 2Rnnq−1, where Rn → ∞, the sum of type I and type II error probabilities approaches at least 1 as n → ∞. To prove the soft converse, we use a sequence of steps to construct a new identification code with a simpler structure which relates to a set system, and then use a lower bound on the normalized maximum pairwise intersection of a set system. To prove the strong converse, we use results on approximation of distributions. The achievability and converse results are generalized to the case of coding over multiple blocks. We finally study message identification over a q-ary uniform permutation channel in the presence of causal block-wise feedback from the receiver, where the encoder receives an entire n-length received block after the transmission of the block is complete. We show that in the presence of feedback, the maximum number of identifiable messages grows doubly exponentially and we present a two-phase achievability *** Codes 68P30, 94A15 Copyright © 2024, The Authors. All rights reserved.

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

Comparing large language models and human programmers for generating programming code

arXiv

引用

arXiv 2024年

作者： Hou, Wenpin Ji, Zhicheng Department of Biostatistics The Mailman School of Public Health Columbia University New York CityNY United States Department of Biostatistics and Bioinformatics Duke University School of Medicine DurhamNC United States

We systematically evaluated the performance of seven large language models in generating programming code using various prompt strategies, programming languages, and task difficulties. GPT-4 substantially outperforms other large language models, including Gemini Ultra and Claude 2. The coding performance of GPT-4 varies considerably with different prompt strategies. In most LeetCode and GeeksforGeeks coding contests evaluated in this study, GPT-4 employing the optimal prompt strategy outperforms 85 percent of human participants. Additionally, GPT-4 demonstrates strong capabilities in translating code between different programming languages and in learning from past errors. The computational efficiency of the code generated by GPT-4 is comparable to that of human programmers. These results suggest that GPT-4 has the potential to serve as a reliable assistant in programming code generation and software development. © 2024, CC BY-SA.

关键词： coding errors

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：