The recent proliferation of large language models (LLMs) holds the potential to revolutionize healthcare, with strong capabilities in diverse medical tasks. Yet, deploying LLMs in high-stakes healthcare settings requires rigorous verification and validation to understand any potential harm. This paper investigates the reliability and viability of using medical knowledge graphs (KGs) for the automated factuality evaluation of LLM-generated responses. To ground this investigation, we introduce FAITH, a framework designed to systematically probe the strengths and limitations of this KG-based approach. FAITH operates without reference answers by decomposing responses into atomic claims, linking them to a medical KG, and scoring them based on evidence paths. Experiments on diverse medical tasks with human subjective evaluations demonstrate that KG-grounded evaluation achieves considerably higher correlations with clinician judgments and can effectively distinguish LLMs with varying capabilities. It is also robust to textual variances. The inherent explainability of its scoring can further help users understand and mitigate the limitations of current LLMs. We conclude that while limitations exist, leveraging KGs is a prominent direction for automated factuality assessment in healthcare.

We are living in an era of "big literature", where the volume of digital scientific publications is growing exponentially. While offering new opportunities, this also poses challenges for understanding literature landscapes, as traditional manual reviewing is no longer feasible. Recent large language models (LLMs) have shown strong capabilities for literature comprehension, yet they are incapable of offering "comprehensive, objective, open and transparent" views desired by systematic reviews due to their limited context windows and trust issues like hallucinations. Here we present LitChat, an end-to-end, interactive and conversational literature agent that augments LLM agents with data-driven discovery tools to facilitate literature exploration. LitChat automatically interprets user queries, retrieves relevant sources, constructs knowledge graphs, and employs diverse data-mining techniques to generate evidence-based insights addressing user needs. We illustrate the effectiveness of LitChat via a case study on AI4Health, highlighting its capacity to quickly navigate the users through large-scale literature landscape with data-based evidence that is otherwise infeasible with traditional means.

Machine learning models increasingly map biological sequence-fitness landscapes to predict mutational effects. Effective evaluation of these models requires benchmarks curated from empirical data. Despite their impressive scales, existing benchmarks lack topographical information regarding the underlying fitness landscapes, which hampers interpretation and comparison of model performance beyond averaged scores. Here, we introduce GraphFLA, a Python framework that constructs and analyzes fitness landscapes from diverse modalities (DNA, RNA, protein, and beyond.), accommodating datasets up to millions of mutants. GraphFLA calculates 20 biologically relevant features that characterize 4 fundamental aspects of landscape topography. By applying GraphFLA to over 5,300 landscapes from ProteinGym, RNAGym, and CIS-BP, we demonstrate its utility in interpreting and comparing the performance of dozens of fitness prediction models, highlighting factors influencing model accuracy and respective advantages of different models. Additionally, we release 155 combinatorially complete empirical fitness landscapes, encompassing over 2.2 million sequences across various modalities. All the codes and datasets are available at https://github.com/COLA-Laboratory/GraphFLA.

Although anthropomorphic AI technologies nowadays are significantly changing the human-AI interaction, the mechanism of how voice-AI chatbots’ anthropomorphism affects users’ continuance behaviour is unclear. To explore the role of human-AI interaction and privacy concerns in the continuance intention of anthropomorphic voice-AI chatbots, a research model, based on parasocial relationship theory, is developed. The research model was then empirically tested against a cross-sectional survey from 473 voice-AI chatbot users and two-wave longitudinal data collected from 271 voice-AI chatbot users. The Structural Equation Modelling (SEM) results indicate that voice-AI chatbots’ anthropomorphism affects continuance intention via both human-AI interaction fluency and human-AI rapport building. In particular, the impact of voice-AI chatbots’ anthropomorphism on voice-AI chatbots’ continuance intention is fully mediated by human-AI interaction fluency and human-AI rapport building. Moreover, the influence of human-AI interaction fluency on voice-AI chatbots’ continuance intention will decrease and the impact of human-AI rapport building on voice-AI chatbots’ continuance intention will increase when users’ privacy concerns are high. The findings provide new insights into AI chatbot research from a human-AI interaction perspective. Developers of voice AI chatbots could focus on anthropomorphism, interaction and user information collection strategies to increase user continuance intention.

Maintaining consumer satisfaction is crucial for the online knowledge payment industry. However, the roles of knowledge contributors in generating consumer satisfaction are not well understood. This study integrates the stereotype content model and social exchange theory to investigate the relationships between contributor stereotypes and consumer satisfaction, as well as the mediating mechanisms and moderating factors. The survey data was collected from 523 Chinese knowledge consumers via a professional survey platform and analysed by combining partial least squares structural equation modelling (PLS-SEM) and fuzzy-set qualitative comparative analysis (fsQCA). The PLS-SEM findings indicate that both warmth and competence of knowledge contributors positively affect consumer satisfaction, and emotional value has a stronger mediating effect in the impact of contributor warmth, whereas functional value has a stronger mediating effect in the impact of contributor competence. Moreover, the effects of contributor warmth and competence are moderated by consumer gender and motivation, and contributor warmth (vs. competence) is more important for females and consumers with hedonic-dominant motivation, while contributor competence (vs. warmth) is more important for males and consumers with utilitarian-dominant motivation. The fsQCA results reinforce the findings from SEM analysis and provide additional insights regarding the configurations of conditions leading to high consumer satisfaction.

Optimization problems are ubiquitous in real-world engineering scenarios where the goals are to enhance interested aspects such as efficiency, productivity, and profitability. However, solving practical optimization problems could be non-trivial, partly due to the presence of a wide range of noises, including environmental noises, model biases, time-domain variations, measurement uncertainties and many other uncontrolled variables. In this paper, we empirically study the effect of noise range, sample size and resampling type on the solution quality of MOEAs when noise is added to decision variables. Our empirical results, conducted on three commonly used Multi-Objective Optimization Problems (MOEAs), i.e. NSGA-II, MOEA/D and IBEA, demonstrate that noise range has more significant impact on the robustness of optimization algorithms compared to sample size and resampling type. In addition, we introduce the concept of bad point, which is able to illustrate how noise affects the performance of different MOEAs.