Recent advancements in large vision-language models (LVLMs) have led to significant progress in generating natural language descriptions for visual content and thus enhancing various applications. One issue with these powerful models is that they sometimes produce texts that are factually inconsistent with the visual input. While there has been some effort to mitigate such inconsistencies in natural image captioning, the factuality of generated captions for structured document images, such as charts, has not received as much scrutiny, posing a potential threat to information reliability in critical applications. This work delves into the factuality aspect by introducing a comprehensive typology of factual errors in generated chart captions. A large-scale human annotation effort provides insight into the error patterns and frequencies in captions crafted by various chart captioning models, ultimately forming the foundation of a novel dataset, CHOCOLATE. Our analysis reveals that even state-of-the-art models, including GPT-4V, frequently produce captions laced with factual inaccuracies. In response to this challenge, we establish the new task of Chart Caption Factual Error Correction and introduce CHARTVE, a model for visual entailment that outperforms proprietary and open-source LVLMs in evaluating factual consistency. Furthermore, we propose C2TFEC, an interpretable two-stage framework that excels at correcting factual errors. This work inaugurates a new domain in factual error correction for chart captions, presenting a novel evaluation mechanism, and demonstrating an effective approach to ensuring the factuality of generated chart captions.

High-quality conversational datasets are essential for developing AI models that can communicate with users. One way to foster deeper interactions between a chatbot and its user is through personas, aspects of the user's character that provide insights into their personality, motivations, and behaviors. Training Natural Language Processing (NLP) models on a diverse and comprehensive persona-based dataset can lead to conversational models that create a deeper connection with the user, and maintain their engagement. In this paper, we leverage the power of Large Language Models (LLMs) to create a large, high-quality conversational dataset from a seed dataset. We propose a Generator-Critic architecture framework to expand the initial dataset, while improving the quality of its conversations. The Generator is an LLM prompted to output conversations. The Critic consists of a mixture of expert LLMs that control the quality of the generated conversations. These experts select the best generated conversations, which we then use to improve the Generator. We release Synthetic-Persona-Chat, consisting of 20k conversations seeded from Persona-Chat. We evaluate the quality of Synthetic-Persona-Chat and our generation framework on different dimensions through extensive experiments, and observe that the losing rate of Synthetic-Persona-Chat against Persona-Chat during Turing test decreases from 17.2% to 8.8% over three iterations.

Shortcut reasoning is an irrational process of inference, which degrades the robustness of an NLP model. While a number of previous work has tackled the identification of shortcut reasoning, there are still two major limitations: (i) a method for quantifying the severity of the discovered shortcut reasoning is not provided; (ii) certain types of shortcut reasoning may be missed. To address these issues, we propose a novel method for identifying shortcut reasoning. The proposed method quantifies the severity of the shortcut reasoning by leveraging out-of-distribution data and does not make any assumptions about the type of tokens triggering the shortcut reasoning. Our experiments on Natural Language Inference and Sentiment Analysis demonstrate that our framework successfully discovers known and unknown shortcut reasoning in the previous work.

Riveter provides a complete easy-to-use pipeline for analyzing verb connotations associated with entities in text corpora. We prepopulate the package with connotation frames of sentiment, power, and agency, which have demonstrated usefulness for capturing social phenomena, such as gender bias, in a broad range of corpora. For decades, lexical frameworks have been foundational tools in computational social science, digital humanities, and natural language processing, facilitating multifaceted analysis of text corpora. But working with verb-centric lexica specifically requires natural language processing skills, reducing their accessibility to other researchers. By organizing the language processing pipeline, providing complete lexicon scores and visualizations for all entities in a corpus, and providing functionality for users to target specific research questions, Riveter greatly improves the accessibility of verb lexica and can facilitate a broad range of future research.

A single language model (LM), despite aligning well with an average labeler through reinforcement learning from human feedback (RLHF), may not universally suit diverse human preferences. Recent approaches therefore opt for customization by collecting multi-dimensional feedback and creating distinct reward models (RMs) for each dimension (e.g., helpfulness, harmlessness, or honesty). Different LMs can then be optimized for different preferences using multi-objective RLHF (MORLHF) with different reward weightings. Yet, RL fine-tuning is unstable and resource-heavy, especially for MORLHF with diverse and usually conflicting objectives. In this paper, we present Multi-Objective Direct Preference Optimization (MODPO), an RL-free algorithm that extends Direct Preference Optimization (DPO) for multiple alignment objectives with minimal overheads. Essentially, MODPO folds language modeling directly into reward modeling, training LMs as implicit collective reward models (cRMs) that combine all objectives with specific weightings. While theoretically guaranteed to produce the same optimal solutions as MORLHF, MODPO is practically more stable and computationally efficient. Empirical results from safety alignment and long-form question answering confirm that MODPO matches or outperforms existing methods, consistently producing a Pareto front of LMs that cater to diverse preferences with 3 times less computational resources compared to MORLHF.

Spotify is testing an AI-powered feature that creates playlists from text prompts. TikTok user @robdad_ posted a short clip of it, captioned, "I just randomly discovered Spotify's ChatGPT?" For the chosen guinea pigs, the feature is available as an option under Your Library after tapping the plus sign to create a new playlist. The news was reported by TechCrunch, which says it received confirmation from Spotify that it's testing AI playlists. It isn't yet clear if the music streamer plans to launch it publicly. "Turn your ideas into playlists using Al," the feature's in-app description reads in the TikTok video (while noting it's only available in English).

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. Pope Francis warned world leaders on Thursday that uninhibited and reckless development of artificial intelligence could pose a profound risk to humanity. The pope made the statements in an address to the world in honor of the upcoming 57th annual World Day of Peace on Jan. 1. "We rightly rejoice and give thanks for the impressive achievements of science and technology, as a result of which countless ills that formerly plagued human life and caused great suffering have been remedied," the pope said.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. Google," it can be tempting to click your way to self-diagnosis -- but an overload of health information can cause its own set of symptoms. "Cyberchondria," a subset of health anxiety, is described as a condition in which an individual excessively searches for health information online. While cyberchrondria may not start as a physical disease, it can cause intense levels of anxiety and fear that can negatively impact a person's health, according to Dr. Maggie Williams, a family physician in Scottsdale, Arizona, and medical director for MDLIVE Virtual Primary Care. Dr. Marc Siegel, clinical professor of medicine at NYU Langone Medical Center and a Fox News medical contributor, said he and his colleagues used to call the condition "medical students' disease." An overload of health information can cause its own set of symptoms called "cyberchondria," or heightened health anxiety. "When you know a little, but not enough, you imagine you have everything and constantly worry," he told Fox News Digital. Although cyberchondria is not listed in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) as a formal diagnosis, it's thought to be closely related to hypochrondria, a more general heightened anxiety about one's health. In 2014, two U.K. researchers, Eoin McElroy and Mark Shevlin, created a "cyberchrondria severity scale" that measures a person's score across eight areas: compulsion, distress, excessiveness, reassurance seeking and mistrust of medical professionals. As Siegel pointed out, the condition is becoming more common over time. "The invention of the internet and then the perfection of search engines created a global hypochondria, where patients searched to find possible explanations for their symptoms," he said. "The invention of the internet and then the perfection of search engines created a global hypochondria, where patients searched to find possible explanations for their symptoms," a doctor told Fox News Digital. "It especially increased during the pandemic, when dogma abounded and everyone was suddenly an expert," Siegel added. A study published in JIMR Formative Research last year found that COVID-19 caused a spike in the condition in spring 2020, as people experienced higher levels of "cyberchondria-related distress and compulsion during the pandemic." "The invention of the internet and then the perfection of search engines created a global hypochondria, where patients searched to find possible explanations for their symptoms." One user shared experiences with cyberchrondria on Reddit: "I thought that I might see something that will ease my mind, but … it makes it all worse and worse.

Oregon-based Happywhale uses all the data it gets from uploaded photos to help it track whale numbers and movements. It is an example of a growing trend - conservation groups using AI to enable members of the public to identify animals or birds. And in return the organisations get a host of crowdsourced data.

Deep anomaly detection methods have become increasingly popular in recent years, with methods like Stacked Autoencoders, Variational Autoencoders, and Generative Adversarial Networks greatly improving the state-of-the-art. Other methods rely on augmenting classical models (such as the One-Class Support Vector Machine), by learning an appropriate kernel function using Neural Networks. Recent developments in representation learning by self-supervision are proving to be very beneficial in the context of anomaly detection. Inspired by the advancements in anomaly detection using self-supervised learning in the field of computer vision, this thesis aims to develop a method for detecting anomalies by exploiting pretext tasks tailored for text corpora. This approach greatly improves the state-of-the-art on two datasets, 20Newsgroups, and AG News, for both semi-supervised and unsupervised anomaly detection, thus proving the potential for self-supervised anomaly detectors in the field of natural language processing.