The age of social media is rife with memes. Understanding and detecting harmful memes pose a significant challenge due to their implicit meaning that is not explicitly conveyed through the surface text and image. However, existing harmful meme detection approaches only recognize superficial harm-indicative signals in an end-to-end classification manner but ignore in-depth cognition of the meme text and image. In this paper, we attempt to detect harmful memes based on advanced reasoning over the interplay of multimodal information in memes. Inspired by the success of Large Language Models (LLMs) on complex reasoning, we first conduct abductive reasoning with LLMs. Then we propose a novel generative framework to learn reasonable thoughts from LLMs for better multimodal fusion and lightweight fine-tuning, which consists of two training stages: 1) Distill multimodal reasoning knowledge from LLMs; and 2) Fine-tune the generative framework to infer harmfulness. Extensive experiments conducted on three meme datasets demonstrate that our proposed approach achieves superior performance than state-of-the-art methods on the harmful meme detection task.

Large multimodal models (LMMs) have evolved from large language models (LLMs) to integrate multiple input modalities, such as visual inputs. This integration augments the capacity of LLMs for tasks requiring visual comprehension and reasoning. However, the extent and limitations of their enhanced abilities are not fully understood, especially when it comes to real-world tasks. To address this gap, we introduce GlitchBench, a novel benchmark derived from video game quality assurance tasks, to test and evaluate the reasoning capabilities of LMMs. Our benchmark is curated from a variety of unusual and glitched scenarios from video games and aims to challenge both the visual and linguistic reasoning powers of LMMs in detecting and interpreting out-of-the-ordinary events. We evaluate multiple state-of-the-art LMMs, and we show that GlitchBench presents a new challenge for these models. Code and data are available at: https://glitchbench.github.io/

Pre-trained models of source code have recently been successfully applied to a wide variety of Software Engineering tasks; they have also seen some practical adoption in practice, e.g. for code completion. Yet, we still know very little about what these pre-trained models learn about source code. In this article, we use probing--simple diagnostic tasks that do not further train the models--to discover to what extent pre-trained models learn about specific aspects of source code. We use an extensible framework to define 15 probing tasks that exercise surface, syntactic, structural and semantic characteristics of source code. We probe 8 pre-trained source code models, as well as a natural language model (BERT) as our baseline. We find that models that incorporate some structural information (such as GraphCodeBERT) have a better representation of source code characteristics. Surprisingly, we find that for some probing tasks, BERT is competitive with the source code models, indicating that there are ample opportunities to improve source-code specific pre-training on the respective code characteristics. We encourage other researchers to evaluate their models with our probing task suite, so that they may peer into the hidden layers of the models and identify what intrinsic code characteristics are encoded.

In recent years, pre-trained large language models (LLMs) have achieved tremendous success in the field of Natural Language Processing (NLP). Prior studies have primarily focused on general and generic domains, with relatively less research on specialized LLMs in the medical field. The specialization and high accuracy requirements for diagnosis in the medical field, as well as the challenges in collecting large-scale data, have constrained the application and development of LLMs in medical scenarios. In the field of ophthalmology, clinical diagnosis mainly relies on doctors' interpretation of reports and making diagnostic decisions. In order to take advantage of LLMs to provide decision support for doctors, we collected three modalities of ophthalmic report data and fine-tuned the LLaMA2 model, successfully constructing an LLM termed the "Ophtha-LLaMA2" specifically tailored for ophthalmic disease diagnosis. Inference test results show that even with a smaller fine-tuning dataset, Ophtha-LLaMA2 performs significantly better in ophthalmic diagnosis compared to other LLMs. It demonstrates that the Ophtha-LLaMA2 exhibits satisfying accuracy and efficiency in ophthalmic disease diagnosis, making it a valuable tool for ophthalmologists to provide improved diagnostic support for patients. This research provides a useful reference for the application of LLMs in the field of ophthalmology, while showcasing the immense potential and prospects in this domain.

In today's technologically driven world, the spread of fake news, particularly during crucial events such as elections, presents an increasing challenge to the integrity of information. To address this challenge, we introduce FakeWatch ElectionShield, an innovative framework carefully designed to detect fake news. We have created a novel dataset of North American election-related news articles through a blend of advanced language models (LMs) and thorough human verification, for precision and relevance. We propose a model hub of LMs for identifying fake news. Our goal is to provide the research community with adaptable and accurate classification models in recognizing the dynamic nature of misinformation. Extensive evaluation of fake news classifiers on our dataset and a benchmark dataset shows our that while state-of-the-art LMs slightly outperform the traditional ML models, classical models are still competitive with their balance of accuracy, explainability, and computational efficiency. This research sets the foundation for future studies to address misinformation related to elections.

In spite of the potential for ground-breaking achievements offered by large language models (LLMs) (e.g., GPT-3), they still lag significantly behind fully-supervised baselines (e.g., fine-tuned BERT) in relation extraction (RE). This is due to the two major shortcomings of LLMs in RE: (1) low relevance regarding entity and relation in retrieved demonstrations for in-context learning; and (2) the strong inclination to wrongly classify NULL examples into other pre-defined labels. In this paper, we propose GPT-RE to bridge the gap between LLMs and fully-supervised baselines. GPT-RE successfully addresses the aforementioned issues by (1) incorporating task-specific entity representations in demonstration retrieval; and (2) enriching the demonstrations with gold label-induced reasoning logic. We evaluate GPT-RE on four widely-used RE datasets, and observe that GPT-RE achieves improvements over not only existing GPT-3 baselines, but also fully-supervised baselines. Specifically, GPT-RE achieves SOTA performances on the Semeval and SciERC datasets, and competitive performances on the TACRED and ACE05 datasets.

The release of ChatGPT, Bard, and other large language model (LLM)-based chatbots has drawn huge attention on foundations models (FMs) worldwide. FMs are massive artificial intelligence (AI) models that are pre-trained on vast amounts of broad data and can be adapted to perform a wide variety of tasks [1]. With numerous projects already underway to explore their potential, it is widely predicted that FMs will serve as the fundamental building blocks for most future AI and artificial generative intelligence (AGI) systems. Many reusable solutions have been proposed to tackle various challenges in designing FM-based systems. However, there is a lack of systematic guidance on the architecture design of FM-based systems. The impact of integrating FMs into software architecture are not fully studied yet. Additionally, the FM's growing capabilities can eventually absorb the other components of AI systems, introducing the moving boundary and interface evolution challenges in architecture design. On the other hand, there are unique challenges on building responsible AI into the architecture of FM-based systems. First, accountability becomes more complex due to the involvement of multiple stakeholders.

Commanding a drone with a natural language is not only user-friendly but also opens the door for emerging language agents to control the drone. Emerging large language models (LLMs) provide a previously impossible opportunity to automatically translate a task description in a natural language to a program that can be executed by the drone. However, powerful LLMs and their vision counterparts are limited in three important ways. First, they are only available as cloud-based services. Sending images to the cloud raises privacy concerns. Second, they are expensive, costing proportionally to the request size. Finally, without expensive fine-tuning, existing LLMs are quite limited in their capability of writing a program for specialized systems like drones. In this paper, we present a system called TypeFly that tackles the above three problems using a combination of edge-based vision intelligence, novel programming language design, and prompt engineering. Instead of the familiar Python, TypeFly gets a cloud-based LLM service to write a program in a small, custom language called MiniSpec, based on task and scene descriptions in English. Such MiniSpec programs are not only succinct (and therefore efficient) but also able to consult the LLM during their execution using a special skill called query. Using a set of increasingly challenging drone tasks, we show that design choices made by TypeFly can reduce both the cost of LLM service and the task execution time by more than 2x. More importantly, query and prompt engineering techniques contributed by TypeFly significantly improve the chance of success of complex tasks.

Foundation models are a current focus of attention in both industry and academia. While they have shown their capabilities in a variety of tasks, in-depth research is required to determine their robustness to distribution shift when used as a basis for supervised machine learning. This is especially important in the context of clinical data, with particular limitations related to data accessibility, lack of pretraining materials, and limited availability of high-quality annotations. In this work, we examine the stability of models based on representations from foundation models under distribution shift. We focus on confounding by provenance, a form of distribution shift that emerges in the context of multi-institutional datasets when there are differences in source-specific language use and class distributions. Using a sampling strategy that synthetically induces varying degrees of distribution shift, we evaluate the extent to which representations from foundation models result in predictions that are inherently robust to confounding by provenance. Additionally, we examine the effectiveness of a straightforward confounding adjustment method inspired by Pearl's conception of backdoor adjustment. Results indicate that while foundation models do show some out-of-the-box robustness to confounding-by-provenance related distribution shifts, this can be considerably improved through adjustment. These findings suggest a need for deliberate adjustment of predictive models using representations from foundation models in the context of source-specific distributional differences.

Model extraction (ME) attacks represent one major threat to Machine-Learning-as-a-Service (MLaaS) platforms by ``stealing'' the functionality of confidential machine-learning models through querying black-box APIs. Over seven years have passed since ME attacks were first conceptualized in the seminal work. During this period, substantial advances have been made in both ME attacks and MLaaS platforms, raising the intriguing question: How has the vulnerability of MLaaS platforms to ME attacks been evolving? In this work, we conduct an in-depth study to answer this critical question. Specifically, we characterize the vulnerability of current, mainstream MLaaS platforms to ME attacks from multiple perspectives including attack strategies, learning techniques, surrogate-model design, and benchmark tasks. Many of our findings challenge previously reported results, suggesting emerging patterns of ME vulnerability. Further, by analyzing the vulnerability of the same MLaaS platforms using historical datasets from the past four years, we retrospectively characterize the evolution of ME vulnerability over time, leading to a set of interesting findings. Finally, we make suggestions about improving the current practice of MLaaS in terms of attack robustness. Our study sheds light on the current state of ME vulnerability in the wild and points to several promising directions for future research.