Large language models (LLMs) have demonstrated an impressive ability to generate codes on competitive programming tasks. However, with limited sample numbers, LLMs still suffer from poor accuracy. Inspired by the process of human programming, we propose a generate-and-edit approach named Self-Edit that utilizes execution results of the generated code from LLMs to improve the code quality on the competitive programming task. We execute the generated code on the example test case provided in the question and wrap execution results into a supplementary comment. Utilizing this comment as guidance, our fault-aware code editor is employed to correct errors in the generated code. We perform extensive evaluations across two competitive programming datasets with nine different LLMs. Compared to directly generating from LLMs, our approach can improve the average of pass@1 by 89\% on APPS-dev, 31\% on APPS-test, and 48\% on HumanEval over nine popular code generation LLMs with parameter sizes ranging from 110M to 175B. Compared to other post-processing methods, our method demonstrates superior accuracy and efficiency.
相關內容
Researchers have invested considerable effort into ensuring that large language models (LLMs) align with human values, using various training techniques, such as instruction tuning and Reinforcement Learning from Human or AI Feedback (RLHF/RLAIF), to guard against text unsafety. However, these defenses remain incredibly vulnerable to some jailbreak attacks, which can cause the model to become overly defensive to sensitive topics or still generate harmful content, leaving the model performance particularly fragile. Therefore, to comprehensively study text safety and output robustness, we propose a latent jailbreak prompt dataset, each involving malicious instruction embedding. Specifically, we instruct the model to complete a regular task, such as translation, where the text to be translated contains malicious instructions. To further analyze the safety and robustness, we design a hierarchical annotation framework. We present a systematic analysis of the safety and robustness of LLMs concerning the position of explicit normal instructions, word replacement (verbs in explicit normal instructions, target groups in malicious instructions, cue words in malicious instructions), and instruction replacement (different explicit normal instructions). Our results show that current LLMs not only have a preference for certain instruction verbs, but also exhibit different jailbreak rates for different instruction verbs in explicit normal instructions. In other words, the probability of generating unsafe content by the model will be reinforced to varying degrees depending on the instruction verb in explicit normal instructions. Code and data are available at //github.com/qiuhuachuan/latent-jailbreak.
In recent years, the use of automated source code generation utilizing transformer-based generative models has expanded, and these models can generate functional code according to the requirements of the developers. However, recent research revealed that these automatically generated source codes can contain vulnerabilities and other quality issues. Despite researchers' and practitioners' attempts to enhance code generation models, retraining and fine-tuning large language models is time-consuming and resource-intensive. Thus, we describe FRANC, a lightweight framework for recommending more secure and high-quality source code derived from transformer-based code generation models. FRANC includes a static filter to make the generated code compilable with heuristics and a quality-aware ranker to sort the code snippets based on a quality score. Moreover, the framework uses prompt engineering to fix persistent quality issues. We evaluated the framework with five Python and Java code generation models and six prompt datasets, including a newly created one in this work (SOEval). The static filter improves 9% to 46% Java suggestions and 10% to 43% Python suggestions regarding compilability. The average improvement over the NDCG@10 score for the ranking system is 0.0763, and the repairing techniques repair the highest 80% of prompts. FRANC takes, on average, 1.98 seconds for Java; for Python, it takes 0.08 seconds.
Developers often face challenges in code understanding, which is crucial for building and maintaining high-quality software systems. Code comments and documentation can provide some context for the code, but are often scarce or missing. This challenge has become even more pressing with the rise of large language model (LLM) based code generation tools. To understand unfamiliar code, most software developers rely on general-purpose search engines to search through various programming information resources, which often requires multiple iterations of query rewriting and information foraging. More recently, developers have turned to online chatbots powered by LLMs, such as ChatGPT, which can provide more customized responses but also incur more overhead as developers need to communicate a significant amount of context to the LLM via a textual interface. In this study, we provide the investigation of an LLM-based conversational UI in the IDE. We aim to understand the promises and obstacles for tools powered by LLMs that are contextually aware, in that they automatically leverage the developer's programming context to answer queries. To this end, we develop an IDE Plugin that allows users to query back-ends such as OpenAI's GPT-3.5 and GPT-4 with high-level requests, like: explaining a highlighted section of code, explaining key domain-specific terms, or providing usage examples for an API. We conduct an exploratory user study with 32 participants to understand the usefulness and effectiveness, as well as individual preferences in the usage of, this LLM-powered information support tool. The study confirms that this approach can aid code understanding more effectively than web search, but the degree of the benefit differed by participants' experience levels.
Large language models (LLMs) have achieved remarkable success in the field of natural language processing, enabling better human-computer interaction using natural language. However, the seamless integration of speech signals into LLMs has not been explored well. The "decoder-only" architecture has also not been well studied for speech processing tasks. In this research, we introduce Speech-LLaMA, a novel approach that effectively incorporates acoustic information into text-based large language models. Our method leverages Connectionist Temporal Classification and a simple audio encoder to map the compressed acoustic features to the continuous semantic space of the LLM. In addition, we further probe the decoder-only architecture for speech-to-text tasks by training a smaller scale randomly initialized speech-LLaMA model from speech-text paired data alone. We conduct experiments on multilingual speech-to-text translation tasks and demonstrate a significant improvement over strong baselines, highlighting the potential advantages of decoder-only models for speech-to-text conversion.
Multilingual speech recognition for both monolingual and code-switching speech is a challenging task. Recently, based on the Mixture of Experts (MoE), many works have made good progress in multilingual and code-switching ASR, but present huge computational complexity with the increase of supported languages. In this work, we propose a computation-efficient network named Language-Routing Mixture of Experts (LR-MoE) for multilingual and code-switching ASR. LR-MoE extracts language-specific representations through the Mixture of Language Experts (MLE), which is guided to learn by a frame-wise language routing mechanism. The weight-shared frame-level language identification (LID) network is jointly trained as the shared pre-router of each MoE layer. Experiments show that the proposed method significantly improves multilingual and code-switching speech recognition performances over baseline with comparable computational efficiency.
Regulation of Multi-Agent Systems (MAS) and Declarative Electronic Institutions (DEIs) was a multidisciplinary research topic of the past decade involving (Physical and Software) Agents and Law since the beginning, but recently evolved towards News-claimed Robot Lawyer since 2016. One of these first proposals of restricting the behaviour of Software Agentswas Electronic Institutions.However, with the recent reformulation of Artificial Neural Networks (ANNs) as Deep Learning (DL), Security, Privacy,Ethical and Legal issues regarding the use of DL has raised concerns in the Artificial Intelligence (AI) Community. Now that the Regulation of MAS is almost correctly addressed, we propose the Regulation of Artificial Neural Networks as Agent-based Training of a special type of regulated Artificial Neural Network that we call Institutional Neural Network (INN).The main purpose of this paper is to bring attention to Artificial Teaching (AT) and to give a tentative answer showing a proof-of-concept implementation of Regulated Deep Learning (RDL). This paper introduces the former concept and provide sI, a language previously used to model declaratively and extend Electronic Institutions, as a means to regulate the execution of Artificial Neural Networks and their interactions with Artificial Teachers (ATs)
Before deploying a language model (LM) within a given domain, it is important to measure its tendency to generate factually incorrect information in that domain. Existing factual generation evaluation methods focus on facts sampled from the LM itself, and thus do not control the set of evaluated facts and might under-represent rare and unlikely facts. We propose FACTOR: Factual Assessment via Corpus TransfORmation, a scalable approach for evaluating LM factuality. FACTOR automatically transforms a factual corpus of interest into a benchmark evaluating an LM's propensity to generate true facts from the corpus vs. similar but incorrect statements. We use our framework to create two benchmarks: Wiki-FACTOR and News-FACTOR. We show that: (i) our benchmark scores increase with model size and improve when the LM is augmented with retrieval; (ii) benchmark score correlates with perplexity, but the two metrics do not always agree on model ranking; and (iii) when perplexity and benchmark score disagree, the latter better reflects factuality in open-ended generation, as measured by human annotators. We make our data and code publicly available in //github.com/AI21Labs/factor.
Existing evaluation metrics for natural language generation (NLG) tasks face the challenges on generalization ability and interpretability. Specifically, most of the well-performed metrics are required to train on evaluation datasets of specific NLG tasks and evaluation dimensions, which may cause over-fitting to task-specific datasets. Furthermore, existing metrics only provide an evaluation score for each dimension without revealing the evidence to interpret how this score is obtained. To deal with these challenges, we propose a simple yet effective metric called DecompEval. This metric formulates NLG evaluation as an instruction-style question answering task and utilizes instruction-tuned pre-trained language models (PLMs) without training on evaluation datasets, aiming to enhance the generalization ability. To make the evaluation process more interpretable, we decompose our devised instruction-style question about the quality of generated texts into the subquestions that measure the quality of each sentence. The subquestions with their answers generated by PLMs are then recomposed as evidence to obtain the evaluation result. Experimental results show that DecompEval achieves state-of-the-art performance in untrained metrics for evaluating text summarization and dialogue generation, which also exhibits strong dimension-level / task-level generalization ability and interpretability.
Bayesian model comparison (BMC) offers a principled approach for assessing the relative merits of competing computational models and propagating uncertainty into model selection decisions. However, BMC is often intractable for the popular class of hierarchical models due to their high-dimensional nested parameter structure. To address this intractability, we propose a deep learning method for performing BMC on any set of hierarchical models which can be instantiated as probabilistic programs. Since our method enables amortized inference, it allows efficient re-estimation of posterior model probabilities and fast performance validation prior to any real-data application. In a series of extensive validation studies, we benchmark the performance of our method against the state-of-the-art bridge sampling method and demonstrate excellent amortized inference across all BMC settings. We then showcase our method by comparing four hierarchical evidence accumulation models that have previously been deemed intractable for BMC due to partly implicit likelihoods. In this application, we corroborate evidence for the recently proposed L\'evy flight model of decision-making and show how transfer learning can be leveraged to enhance training efficiency. We provide reproducible code for all analyses and an open-source implementation of our method.
Since the 1950s, machine translation (MT) has become one of the important tasks of AI and development, and has experienced several different periods and stages of development, including rule-based methods, statistical methods, and recently proposed neural network-based learning methods. Accompanying these staged leaps is the evaluation research and development of MT, especially the important role of evaluation methods in statistical translation and neural translation research. The evaluation task of MT is not only to evaluate the quality of machine translation, but also to give timely feedback to machine translation researchers on the problems existing in machine translation itself, how to improve and how to optimise. In some practical application fields, such as in the absence of reference translations, the quality estimation of machine translation plays an important role as an indicator to reveal the credibility of automatically translated target languages. This report mainly includes the following contents: a brief history of machine translation evaluation (MTE), the classification of research methods on MTE, and the the cutting-edge progress, including human evaluation, automatic evaluation, and evaluation of evaluation methods (meta-evaluation). Manual evaluation and automatic evaluation include reference-translation based and reference-translation independent participation; automatic evaluation methods include traditional n-gram string matching, models applying syntax and semantics, and deep learning models; evaluation of evaluation methods includes estimating the credibility of human evaluations, the reliability of the automatic evaluation, the reliability of the test set, etc. Advances in cutting-edge evaluation methods include task-based evaluation, using pre-trained language models based on big data, and lightweight optimisation models using distillation techniques.
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