Very large language models (LLMs) perform extremely well on a spectrum of NLP tasks in a zero-shot setting. However, little is known about their performance on human-level NLP problems which rely on understanding psychological concepts, such as assessing personality traits. In this work, we investigate the zero-shot ability of GPT-3 to estimate the Big 5 personality traits from users' social media posts. Through a set of systematic experiments, we find that zero-shot GPT-3 performance is somewhat close to an existing pre-trained SotA for broad classification upon injecting knowledge about the trait in the prompts. However, when prompted to provide fine-grained classification, its performance drops to close to a simple most frequent class (MFC) baseline. We further analyze where GPT-3 performs better, as well as worse, than a pretrained lexical model, illustrating systematic errors that suggest ways to improve LLMs on human-level NLP tasks.
相關內容
We study Bayesian optimization (BO) in high-dimensional and non-stationary scenarios. Existing algorithms for such scenarios typically require extensive hyperparameter tuning, which limits their practical effectiveness. We propose a framework, called BALLET, which adaptively filters for a high-confidence region of interest (ROI) as a superlevel-set of a nonparametric probabilistic model such as a Gaussian process (GP). Our approach is easy to tune, and is able to focus on local region of the optimization space that can be tackled by existing BO methods. The key idea is to use two probabilistic models: a coarse GP to identify the ROI, and a localized GP for optimization within the ROI. We show theoretically that BALLET can efficiently shrink the search space, and can exhibit a tighter regret bound than standard BO without ROI filtering. We demonstrate empirically the effectiveness of BALLET on both synthetic and real-world optimization tasks.
Depression is the most prevalent and serious mental illness, which induces grave financial and societal ramifications. Depression detection is key for early intervention to mitigate those consequences. Such a high-stake decision inherently necessitates interpretability. Although a few depression detection studies attempt to explain the decision based on the importance score or attention weights, these explanations misalign with the clinical depression diagnosis criterion that is based on depressive symptoms. To fill this gap, we follow the computational design science paradigm to develop a novel Multi-Scale Temporal Prototype Network (MSTPNet). MSTPNet innovatively detects and interprets depressive symptoms as well as how long they last. Extensive empirical analyses using a large-scale dataset show that MSTPNet outperforms state-of-the-art depression detection methods with an F1-score of 0.851. This result also reveals new symptoms that are unnoted in the survey approach, such as sharing admiration for a different life. We further conduct a user study to demonstrate its superiority over the benchmarks in interpretability. This study contributes to IS literature with a novel interpretable deep learning model for depression detection in social media. In practice, our proposed method can be implemented in social media platforms to provide personalized online resources for detected depressed patients.
Modern language models capture a large body of factual knowledge. However, some facts can be incorrectly induced or become obsolete over time, resulting in factually incorrect generations. This has led to the development of various editing methods that allow updating facts encoded by the model. Evaluation of these methods has primarily focused on testing whether an individual fact has been successfully injected, and if similar predictions for other subjects have not changed. Here we argue that such evaluation is limited, since injecting one fact (e.g. ``Jack Depp is the son of Johnny Depp'') introduces a ``ripple effect'' in the form of additional facts that the model needs to update (e.g.``Jack Depp is the sibling of Lily-Rose Depp''). To address this issue, we propose a novel set of evaluation criteria that consider the implications of an edit on related facts. Using these criteria, we then construct \ripple{}, a diagnostic benchmark of 5K factual edits, capturing a variety of types of ripple effects. We evaluate prominent editing methods on \ripple{}, showing that current methods fail to introduce consistent changes in the model's knowledge. In addition, we find that a simple in-context editing baseline obtains the best scores on our benchmark, suggesting a promising research direction for model editing.
The zero-shot learning capabilities of large language models (LLMs) make them ideal for text classification without annotation or supervised training. Many studies have shown impressive results across multiple tasks. While tasks, data, and results differ widely, their similarities to human annotation can aid us in tackling new tasks with minimal expenses. We evaluate using 5 state-of-the-art LLMs as "annotators" on 5 different tasks (age, gender, topic, sentiment prediction, and hate speech detection), across 4 languages: English, French, German, and Spanish. No single model excels at all tasks, across languages, or across all labels within a task. However, aggregation techniques designed for human annotators perform substantially better than any one individual model. Overall, though, LLMs do not rival even simple supervised models, so they do not (yet) replace the need for human annotation. We also discuss the tradeoffs between speed, accuracy, cost, and bias when it comes to aggregated model labeling versus human annotation.
We assessed the performance of commercial Large Language Models (LLMs) GPT-3.5-Turbo and GPT-4 on tasks from the 2023 BioASQ challenge. In Task 11b Phase B, which is focused on answer generation, both models demonstrated competitive abilities with leading systems. Remarkably, they achieved this with simple zero-shot learning, grounded with relevant snippets. Even without relevant snippets, their performance was decent, though not on par with the best systems. Interestingly, the older and cheaper GPT-3.5-Turbo system was able to compete with GPT-4 in the grounded Q&A setting on factoid and list answers. In Task 11b Phase A, focusing on retrieval, query expansion through zero-shot learning improved performance, but the models fell short compared to other systems. The code needed to rerun these experiments is available through GitHub.
Fuzz drivers are a necessary component of API fuzzing. However, automatically generating correct and robust fuzz drivers is a difficult task. Compared to existing approaches, LLM-based (Large Language Model) generation is a promising direction due to its ability to operate with low requirements on consumer programs, leverage multiple dimensions of API usage information, and generate human-friendly output code. Nonetheless, the challenges and effectiveness of LLM-based fuzz driver generation remain unclear. To address this, we conducted a study on the effects, challenges, and techniques of LLM-based fuzz driver generation. Our study involved building a quiz with 86 fuzz driver generation questions from 30 popular C projects, constructing precise effectiveness validation criteria for each question, and developing a framework for semi-automated evaluation. We designed five query strategies, evaluated 36,506 generated fuzz drivers. Furthermore, the drivers were compared with manually written ones to obtain practical insights. Our evaluation revealed that: while the overall performance was promising (passing 91% of questions), there were still practical challenges in filtering out the ineffective fuzz drivers for large scale application; basic strategies achieved a decent correctness rate (53%), but struggled with complex API-specific usage questions. In such cases, example code snippets and iterative queries proved helpful; while LLM-generated drivers showed competent fuzzing outcomes compared to manually written ones, there was still significant room for improvement, such as incorporating semantic oracles for logical bugs detection.
Adam is a commonly used stochastic optimization algorithm in machine learning. However, its convergence is still not fully understood, especially in the non-convex setting. This paper focuses on exploring hyperparameter settings for the convergence of vanilla Adam and tackling the challenges of non-ergodic convergence related to practical application. The primary contributions are summarized as follows: firstly, we introduce precise definitions of ergodic and non-ergodic convergence, which cover nearly all forms of convergence for stochastic optimization algorithms. Meanwhile, we emphasize the superiority of non-ergodic convergence over ergodic convergence. Secondly, we establish a weaker sufficient condition for the ergodic convergence guarantee of Adam, allowing a more relaxed choice of hyperparameters. On this basis, we achieve the almost sure ergodic convergence rate of Adam, which is arbitrarily close to $o(1/\sqrt{K})$. More importantly, we prove, for the first time, that the last iterate of Adam converges to a stationary point for non-convex objectives. Finally, we obtain the non-ergodic convergence rate of $O(1/K)$ for function values under the Polyak-Lojasiewicz (PL) condition. These findings build a solid theoretical foundation for Adam to solve non-convex stochastic optimization problems.
This paper presents a comprehensive and practical guide for practitioners and end-users working with Large Language Models (LLMs) in their downstream natural language processing (NLP) tasks. We provide discussions and insights into the usage of LLMs from the perspectives of models, data, and downstream tasks. Firstly, we offer an introduction and brief summary of current GPT- and BERT-style LLMs. Then, we discuss the influence of pre-training data, training data, and test data. Most importantly, we provide a detailed discussion about the use and non-use cases of large language models for various natural language processing tasks, such as knowledge-intensive tasks, traditional natural language understanding tasks, natural language generation tasks, emergent abilities, and considerations for specific tasks.We present various use cases and non-use cases to illustrate the practical applications and limitations of LLMs in real-world scenarios. We also try to understand the importance of data and the specific challenges associated with each NLP task. Furthermore, we explore the impact of spurious biases on LLMs and delve into other essential considerations, such as efficiency, cost, and latency, to ensure a comprehensive understanding of deploying LLMs in practice. This comprehensive guide aims to provide researchers and practitioners with valuable insights and best practices for working with LLMs, thereby enabling the successful implementation of these models in a wide range of NLP tasks. A curated list of practical guide resources of LLMs, regularly updated, can be found at \url{//github.com/Mooler0410/LLMsPracticalGuide}.
Knowledge enhanced pre-trained language models (K-PLMs) are shown to be effective for many public tasks in the literature but few of them have been successfully applied in practice. To address this problem, we propose K-AID, a systematic approach that includes a low-cost knowledge acquisition process for acquiring domain knowledge, an effective knowledge infusion module for improving model performance, and a knowledge distillation component for reducing the model size and deploying K-PLMs on resource-restricted devices (e.g., CPU) for real-world application. Importantly, instead of capturing entity knowledge like the majority of existing K-PLMs, our approach captures relational knowledge, which contributes to better-improving sentence-level text classification and text matching tasks that play a key role in question answering (QA). We conducted a set of experiments on five text classification tasks and three text matching tasks from three domains, namely E-commerce, Government, and Film&TV, and performed online A/B tests in E-commerce. Experimental results show that our approach is able to achieve substantial improvement on sentence-level question answering tasks and bring beneficial business value in industrial settings.
This paper surveys and organizes research works in a new paradigm in natural language processing, which we dub "prompt-based learning". Unlike traditional supervised learning, which trains a model to take in an input x and predict an output y as P(y|x), prompt-based learning is based on language models that model the probability of text directly. To use these models to perform prediction tasks, the original input x is modified using a template into a textual string prompt x' that has some unfilled slots, and then the language model is used to probabilistically fill the unfilled information to obtain a final string x, from which the final output y can be derived. This framework is powerful and attractive for a number of reasons: it allows the language model to be pre-trained on massive amounts of raw text, and by defining a new prompting function the model is able to perform few-shot or even zero-shot learning, adapting to new scenarios with few or no labeled data. In this paper we introduce the basics of this promising paradigm, describe a unified set of mathematical notations that can cover a wide variety of existing work, and organize existing work along several dimensions, e.g.the choice of pre-trained models, prompts, and tuning strategies. To make the field more accessible to interested beginners, we not only make a systematic review of existing works and a highly structured typology of prompt-based concepts, but also release other resources, e.g., a website //pretrain.nlpedia.ai/ including constantly-updated survey, and paperlist.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191