Large language models (LLMs) have accomplished remarkable reasoning performance in various domains. However, in the domain of reasoning tasks, we discover a frailty: LLMs are surprisingly brittle to the ordering of the premises, despite the fact that such ordering does not alter the underlying task. In particular, we observe that LLMs achieve the best performance when the premise order aligns with the context required in intermediate reasoning steps. For example, in deductive reasoning tasks, presenting the premises in the same order as the ground truth proof in the prompt (as opposed to random ordering) drastically increases the model's accuracy. We first examine the effect of premise ordering on deductive reasoning on a variety of LLMs, and our evaluation shows that permuting the premise order can cause a performance drop of over 30%. In addition, we release the benchmark R-GSM, based on GSM8K, to examine the ordering effect for mathematical problem-solving, and we again observe a significant drop in accuracy, relative to the original GSM8K benchmark.
相關內容
Stakeholders often describe system requirements using natural language which are then converted to formal syntax by a domain-expert leading to increased design costs. This paper assesses the capabilities of Large Language Models (LLMs) in converting between natural language descriptions and formal specifications. Existing work has evaluated the capabilities of LLMs in generating formal syntax such as source code but such experiments are typically hand-crafted and use problems that are likely to be in the training set of LLMs, and often require human-annotated datasets. We propose an approach that can use two copies of an LLM in conjunction with an off-the-shelf verifier to automatically evaluate its translation abilities without any additional human input. Our approach generates formal syntax using language grammars to automatically generate a dataset. We conduct an empirical evaluation to measure the accuracy of this translation task and show that SOTA LLMs cannot adequately solve this task, limiting their current utility in the design of complex systems.
Pretrained language models (PLMs) have made remarkable progress in table-to-text generation tasks. However, the lack of domain-specific knowledge makes it challenging to bridge the topological gap between tabular data and text, especially in real-world applications with limited resources. To mitigate the limitation of insufficient labeled data, we propose a novel framework: Adapt-Knowledge-to-Generate (AKG). The core insight of AKG is to adapt unlabeled domain-specific knowledge into the model, which brings at least three benefits: (1) it injects representation of normal table-related descriptions to bridge the topological gap between tabular data and texts; (2) it enables us to use large amounts of unlabeled domain-specific knowledge fully, which can alleviate the PLMs' inherent shortcomings of lacking domain knowledge; (3) it allows us to design various tasks to employ the domain-specific knowledge. Extensive experiments and analyses are conducted on three open-domain, few-shot natural language generation (NLG) data sets: Humans, Songs, and Books. Compared to previous state-of-the-art approaches, our model achieves superior performance in terms of both fluency and accuracy as judged by human and automatic evaluations.
Large language models (LLMs) have demonstrated remarkable capabilities across various research domains, including the field of Information Retrieval (IR). However, the responses generated by off-the-shelf LLMs tend to be generic, i.e., cannot capture the distinctiveness of each document with similar content. This limits the performance of LLMs in IR because finding and distinguishing relevant documents from substantial similar documents is a typical problem in many IR tasks. To address this issue, we propose an unsupervised alignment method, namely Reinforcement Learning from Contrastive Feedback (RLCF), empowering LLMs to generate both high-quality and context-specific responses. Our approach constructs unsupervised contrastive feedback signals based on similar document groups, and adopts a reward function, named group-wise reciprocal rank, to optimize LLMs within a standard Proximal Policy Optimization. We conduct extensive experiments to evaluate the effectiveness of RLCF on LLMs built with different languages and parameter sizes on multiple downstream IR applications. RLCF significantly outperforms existing alignment methods, and RLCF-optimized LLMs demonstrate considerable improvement in generating responses with distinctiveness.
Large language models (LLMs) have been increasingly employed for (interactive) decision-making, via the development of LLM-based autonomous agents. Despite their emerging successes, the performance of LLM agents in decision-making has not been fully investigated through quantitative metrics, especially in the multi-agent setting when they interact with each other, a typical scenario in real-world LLM-agent applications. To better understand the limits of LLM agents in these interactive environments, we propose to study their interactions in benchmark decision-making settings in online learning and game theory, through the performance metric of \emph{regret}. We first empirically study the {no-regret} behaviors of LLMs in canonical (non-stationary) online learning problems, as well as the emergence of equilibria when LLM agents interact through playing repeated games. We then provide some theoretical insights into the no-regret behaviors of LLM agents, under certain assumptions on the supervised pre-training and the rationality model of human decision-makers who generate the data. Notably, we also identify (simple) cases where advanced LLMs such as GPT-4 fail to be no-regret. To promote the no-regret behaviors, we propose a novel \emph{unsupervised} training loss of \emph{regret-loss}, which, in contrast to the supervised pre-training loss, does not require the labels of (optimal) actions. We then establish the statistical guarantee of generalization bound for regret-loss minimization, followed by the optimization guarantee that minimizing such a loss may automatically lead to known no-regret learning algorithms. Our further experiments demonstrate the effectiveness of our regret-loss, especially in addressing the above ``regrettable'' cases.
Large language models (LLMs) have shown remarkable progress in automated code generation. Yet, incorporating LLM-based code generation into real-life software projects poses challenges, as the generated code may contain errors in API usage, class, data structure, or missing project-specific information. As much of this project-specific context cannot fit into the prompts of LLMs, we must find ways to allow the model to explore the project-level code context. To this end, this paper puts forward a novel approach, termed ProCoder, which iteratively refines the project-level code context for precise code generation, guided by the compiler feedback. In particular, ProCoder first leverages compiler techniques to identify a mismatch between the generated code and the project's context. It then iteratively aligns and fixes the identified errors using information extracted from the code repository. We integrate ProCoder with two representative LLMs, i.e., GPT-3.5-Turbo and Code Llama (13B), and apply it to Python code generation. Experimental results show that ProCoder significantly improves the vanilla LLMs by over 80% in generating code dependent on project context, and consistently outperforms the existing retrieval-based code generation baselines.
Large language models (LLMs) have demonstrated remarkable capabilities across a wide range of tasks in various domains. Despite their impressive performance, they can be unreliable due to factual errors in their generations. Assessing their confidence and calibrating them across different tasks can help mitigate risks and enable LLMs to produce better generations. There has been a lot of recent research aiming to address this, but there has been no comprehensive overview to organize it and outline the main lessons learned. The present survey aims to bridge this gap. In particular, we outline the challenges and we summarize recent technical advancements for LLM confidence estimation and calibration. We further discuss their applications and suggest promising directions for future work.
Performance of large language models (LLMs) may vary with different prompts or instructions of even the same task. One commonly recognized factor for this phenomenon is the model's familiarity with the given prompt or instruction, which is typically estimated by its perplexity. However, finding the prompt with the lowest perplexity is challenging, given the enormous space of possible prompting phrases. In this paper, we propose monotonic paraphrasing (MonoPara), an end-to-end decoding strategy that paraphrases given prompts or instructions into their lower perplexity counterparts based on an ensemble of a paraphrase LM for prompt (or instruction) rewriting, and a target LM (i.e. the prompt or instruction executor) that constrains the generation for lower perplexity. The ensemble decoding process can efficiently paraphrase the original prompt without altering its semantic meaning, while monotonically decreasing the perplexity of each generation as calculated by the target LM. We explore in detail both greedy and search-based decoding as two alternative decoding schemes of MonoPara. Notably, MonoPara does not require any training and can monotonically lower the perplexity of the paraphrased prompt or instruction, leading to improved performance of zero-shot LM prompting as evaluated on a wide selection of tasks. In addition, MonoPara is also shown to effectively improve LMs' generalization on perturbed and unseen task instructions.
Large vision-language models (VLMs) such as GPT-4 have achieved exceptional performance across various multi-modal tasks. However, the deployment of VLMs necessitates substantial energy consumption and computational resources. Once attackers maliciously induce high energy consumption and latency time (energy-latency cost) during inference of VLMs, it will exhaust computational resources. In this paper, we explore this attack surface about availability of VLMs and aim to induce high energy-latency cost during inference of VLMs. We find that high energy-latency cost during inference of VLMs can be manipulated by maximizing the length of generated sequences. To this end, we propose verbose images, with the goal of crafting an imperceptible perturbation to induce VLMs to generate long sentences during inference. Concretely, we design three loss objectives. First, a loss is proposed to delay the occurrence of end-of-sequence (EOS) token, where EOS token is a signal for VLMs to stop generating further tokens. Moreover, an uncertainty loss and a token diversity loss are proposed to increase the uncertainty over each generated token and the diversity among all tokens of the whole generated sequence, respectively, which can break output dependency at token-level and sequence-level. Furthermore, a temporal weight adjustment algorithm is proposed, which can effectively balance these losses. Extensive experiments demonstrate that our verbose images can increase the length of generated sequences by 7.87 times and 8.56 times compared to original images on MS-COCO and ImageNet datasets, which presents potential challenges for various applications. Our code is available at //github.com/KuofengGao/Verbose_Images.
With the recent development of large language models (LLMs), models that focus on certain domains and languages have been discussed for their necessity. There is also a growing need for benchmarks to evaluate the performance of current LLMs in each domain. Therefore, in this study, we constructed a benchmark comprising multiple tasks specific to the Japanese and financial domains and performed benchmark measurements on some models. Consequently, we confirmed that GPT-4 is currently outstanding, and that the constructed benchmarks function effectively. According to our analysis, our benchmark can differentiate benchmark scores among models in all performance ranges by combining tasks with different difficulties.
Large Language Models (LLMs) have significantly advanced natural language processing (NLP) with their impressive language understanding and generation capabilities. However, their performance may be suboptimal for long-tail or domain-specific tasks due to limited exposure to domain-specific knowledge and vocabulary. Additionally, the lack of transparency of most state-of-the-art (SOTA) LLMs, which can only be accessed via APIs, impedes further fine-tuning with custom data. Moreover, data privacy is a significant concern. To address these challenges, we propose the novel Parametric Knowledge Guiding (PKG) framework, which equips LLMs with a knowledge-guiding module to access relevant knowledge at runtime without altering the LLMs' parameters. Our PKG is based on open-source "white-box" small language models, allowing offline storage of any knowledge that LLMs require. We demonstrate that our PKG framework can enhance the performance of "black-box" LLMs on a range of long-tail and domain-specific downstream tasks requiring factual, tabular, medical, and multimodal knowledge.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191