Are Large Language Models Reliable Judges? A Study on the Factuality Evaluation Capabilities of LLMs
In recent years, Large Language Models (LLMs) have gained immense attention due to their notable emergent capabilities, surpassing those seen in earlier language models. A particularly intriguing application of LLMs is their role as evaluators for texts produced by various generative models. In this study, we delve into the potential of LLMs as reliable assessors of factual consistency in summaries generated by text-generation models. Initially, we introduce an innovative approach for factuality assessment using LLMs. This entails employing a singular LLM for the entirety of the question-answering-based factuality scoring process. Following this, we examine the efficacy of various LLMs in direct factuality scoring, benchmarking them against traditional measures and human annotations. Contrary to initial expectations, our results indicate a lack of significant correlations between factuality metrics and human evaluations, specifically for GPT-4 and PaLM-2. Notable correlations were only observed with GPT-3.5 across two factuality subcategories. These consistent findings across various factual error categories suggest a fundamental limitation in the current LLMs' capability to accurately gauge factuality. This version presents the information more concisely while maintaining the main points and findings of the original text.
相關內容
Cybersecurity challenges and the need for awareness are well-recognized in developed countries, but this still needs attention in less-developed countries. With the expansion of technology, security concerns are also becoming more prevalent worldwide. This paper presents a design and creation research study exploring which factors we should consider when designing cybersecurity awareness solutions for young people in developing countries. We have developed prototypes of mini-cybersecurity awareness applications and conducted a pilot study with eight participants (aged 16-30) from Gambia, Eritrea, and Syria. Our findings show that factors like the influence of culture and social constructs, literacy, and language competence, the way of introducing cybersecurity terms and concepts, and the need for reflection are essential to consider when designing and developing cybersecurity awareness solutions for target users in developing countries. The findings of this study will guide future researchers to design more inclusive cybersecurity awareness solutions for users in developing countries.
Proximal Policy Optimization algorithm employing a clipped surrogate objective (PPO-Clip) is a prominent exemplar of the policy optimization methods. However, despite its remarkable empirical success, PPO-Clip lacks theoretical substantiation to date. In this paper, we contribute to the field by establishing the first global convergence results of a PPO-Clip variant in both tabular and neural function approximation settings. Our findings highlight the $O(1/\sqrt{T})$ min-iterate convergence rate specifically in the context of neural function approximation. We tackle the inherent challenges in analyzing PPO-Clip through three central concepts: (i) We introduce a generalized version of the PPO-Clip objective, illuminated by its connection with the hinge loss. (ii) Employing entropic mirror descent, we establish asymptotic convergence for tabular PPO-Clip with direct policy parameterization. (iii) Inspired by the tabular analysis, we streamline convergence analysis by introducing a two-step policy improvement approach. This decouples policy search from complex neural policy parameterization using a regression-based update scheme. Furthermore, we gain deeper insights into the efficacy of PPO-Clip by interpreting these generalized objectives. Our theoretical findings also mark the first characterization of the influence of the clipping mechanism on PPO-Clip convergence. Importantly, the clipping range affects only the pre-constant of the convergence rate.
Conventional Federated Domain Adaptation (FDA) approaches usually demand an abundance of assumptions, which makes them significantly less feasible for real-world situations and introduces security hazards. This paper relaxes the assumptions from previous FDAs and studies a more practical scenario named Universal Federated Domain Adaptation (UFDA). It only requires the black-box model and the label set information of each source domain, while the label sets of different source domains could be inconsistent, and the target-domain label set is totally blind. Towards a more effective solution for our newly proposed UFDA scenario, we propose a corresponding methodology called Hot-Learning with Contrastive Label Disambiguation (HCLD). It particularly tackles UFDA's domain shifts and category gaps problems by using one-hot outputs from the black-box models of various source domains. Moreover, to better distinguish the shared and unknown classes, we further present a cluster-level strategy named Mutual-Voting Decision (MVD) to extract robust consensus knowledge across peer classes from both source and target domains. Extensive experiments on three benchmark datasets demonstrate that our method achieves comparable performance for our UFDA scenario with much fewer assumptions, compared to previous methodologies with comprehensive additional assumptions.
In recent years, the rapid advancement and impressive capabilities of Large Language Models (LLMs) have been evident across various domains. This paper explores the application, implications, and potential of LLMs in building energy efficiency and decarbonization studies. The wide-ranging capabilities of LLMs are examined in the context of the building energy field, including intelligent control systems, code generation, data infrastructure, knowledge extraction, and education. Despite the promising potential of LLMs, challenges including complex and expensive computation, data privacy, security and copyright, complexity in fine-tuned LLMs, and self-consistency are discussed. The paper concludes with a call for future research focused on the enhancement of LLMs for domain-specific tasks, multi-modal LLMs, and collaborative research between AI and energy experts.
The maturation of cognition, from introspection to understanding others, has long been a hallmark of human development. This position paper posits that for AI systems to truly emulate or approach human-like interactions, especially within multifaceted environments populated with diverse agents, they must first achieve an in-depth and nuanced understanding of self. Drawing parallels with the human developmental trajectory from self-awareness to mentalizing (also called theory of mind), the paper argues that the quality of an autonomous agent's introspective capabilities of self are crucial in mirroring quality human-like understandings of other agents. While counterarguments emphasize practicality, computational efficiency, and ethical concerns, this position proposes a development approach, blending algorithmic considerations of self-referential processing. Ultimately, the vision set forth is not merely of machines that compute but of entities that introspect, empathize, and understand, harmonizing with the complex compositions of human cognition.
Traditional neural networks are simple to train but they typically produce overconfident predictions. In contrast, Bayesian neural networks provide good uncertainty quantification but optimizing them is time consuming due to the large parameter space. This paper proposes to combine the advantages of both approaches by performing Variational Inference in the Final layer Output space (VIFO), because the output space is much smaller than the parameter space. We use neural networks to learn the mean and the variance of the probabilistic output. Like standard, non-Beyesian models, VIFO enjoys simple training and one can use Rademacher complexity to provide risk bounds for the model. On the other hand, using the Bayesian formulation we incorporate collapsed variational inference with VIFO which significantly improves the performance in practice. Experiments show that VIFO and ensembles of VIFO provide a good tradeoff in terms of run time and uncertainty quantification, especially for out of distribution data.
In recent years, significant advancements have been made in the text generation capabilities of Large Language Models (LLMs), demonstrating exceptional performance in downstream tasks such as abstract summarization, dialogue generation, and data-to-text conversion. However, their generative abilities also pose risks such as the rapid spread of fake news, infringement of datasets/LLM copyrights, and challenges to academic integrity. Text watermarking technology emerges as a potential solution. By embedding invisible yet detectable patterns in generated texts, it helps in tracking and verifying text origins, thus preventing misuse and piracy. This survey aims to comprehensively summarize current text watermarking technologies, covering three main aspects: (1) an overview and comparison of different text watermarking techniques; (2) evaluation methods for text watermarking algorithms, including their success rate, impact on text quality, robustness, and unforgeability; (3) potential applications of text watermarking technologies. This survey aims to help researchers thoroughly understanding the text watermarking technologies, thereby fostering further development.
The wayward quality of continuous prompts stresses the importance of their interpretability as unexpected and unpredictable behaviors appear following training, especially in the context of large language models automating people-sensitive tasks such as resume screening. In this paper we present a novel method of constructing continuous prompts via discrete prompt embeddings and evaluate improvements to continuous prompt interpretability and inference accuracy. For a set of manually designed discrete prompts $\mathcal{D}$, which we tokenize each into tensor form, we train a model to predict the weights such that the linear combinations of those prompts correspond to higher performance on natural language understanding tasks.
Graph Neural Networks (GNNs) have been studied from the lens of expressive power and generalization. However, their optimization properties are less well understood. We take the first step towards analyzing GNN training by studying the gradient dynamics of GNNs. First, we analyze linearized GNNs and prove that despite the non-convexity of training, convergence to a global minimum at a linear rate is guaranteed under mild assumptions that we validate on real-world graphs. Second, we study what may affect the GNNs' training speed. Our results show that the training of GNNs is implicitly accelerated by skip connections, more depth, and/or a good label distribution. Empirical results confirm that our theoretical results for linearized GNNs align with the training behavior of nonlinear GNNs. Our results provide the first theoretical support for the success of GNNs with skip connections in terms of optimization, and suggest that deep GNNs with skip connections would be promising in practice.
Graph Neural Networks (GNNs) have recently become increasingly popular due to their ability to learn complex systems of relations or interactions arising in a broad spectrum of problems ranging from biology and particle physics to social networks and recommendation systems. Despite the plethora of different models for deep learning on graphs, few approaches have been proposed thus far for dealing with graphs that present some sort of dynamic nature (e.g. evolving features or connectivity over time). In this paper, we present Temporal Graph Networks (TGNs), a generic, efficient framework for deep learning on dynamic graphs represented as sequences of timed events. Thanks to a novel combination of memory modules and graph-based operators, TGNs are able to significantly outperform previous approaches being at the same time more computationally efficient. We furthermore show that several previous models for learning on dynamic graphs can be cast as specific instances of our framework. We perform a detailed ablation study of different components of our framework and devise the best configuration that achieves state-of-the-art performance on several transductive and inductive prediction tasks for dynamic graphs.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191