Generative multimodal content is increasingly prevalent in much of the content creation arena, as it has the potential to allow artists and media personnel to create pre-production mockups by quickly bringing their ideas to life. The generation of audio from text prompts is an important aspect of such processes in the music and film industry. Many of the recent diffusion-based text-to-audio models focus on training increasingly sophisticated diffusion models on a large set of datasets of prompt-audio pairs. These models do not explicitly focus on the presence of concepts or events and their temporal ordering in the output audio with respect to the input prompt. Our hypothesis is focusing on how these aspects of audio generation could improve audio generation performance in the presence of limited data. As such, in this work, using an existing text-to-audio model Tango, we synthetically create a preference dataset where each prompt has a winner audio output and some loser audio outputs for the diffusion model to learn from. The loser outputs, in theory, have some concepts from the prompt missing or in an incorrect order. We fine-tune the publicly available Tango text-to-audio model using diffusion-DPO (direct preference optimization) loss on our preference dataset and show that it leads to improved audio output over Tango and AudioLDM2, in terms of both automatic- and manual-evaluation metrics.

In the rapidly evolving digital era, there is an increasing demand for concise information as individuals seek to distil key insights from various sources. Recent attention from researchers on Multi-document Summarisation (MDS) has resulted in diverse datasets covering customer reviews, academic papers, medical and legal documents, and news articles. However, the English-centric nature of these datasets has created a conspicuous void for multilingual datasets in today's globalised digital landscape, where linguistic diversity is celebrated. Media platforms such as British Broadcasting Corporation (BBC) have disseminated news in 20+ languages for decades. With only 380 million people speaking English natively as their first language, accounting for less than 5% of the global population, the vast majority primarily relies on other languages. These facts underscore the need for inclusivity in MDS research, utilising resources from diverse languages. Recognising this gap, we present the Multilingual Dataset for Multi-document Summarisation (M2DS), which, to the best of our knowledge, is the first dataset of its kind. It includes document-summary pairs in five languages from BBC articles published during the 2010-2023 period. This paper introduces M2DS, emphasising its unique multilingual aspect, and includes baseline scores from state-of-the-art MDS models evaluated on our dataset.

Query rewriting aims to generate a new query that can complement the original query to improve the information retrieval system. Recent studies on query rewriting, such as query2doc (Q2D), query2expand (Q2E) and querey2cot (Q2C), rely on the internal knowledge of Large Language Models (LLMs) to generate a relevant passage to add information to the query. Nevertheless, the efficacy of these methodologies may markedly decline in instances where the requisite knowledge is not encapsulated within the model's intrinsic parameters. In this paper, we propose a novel structured query rewriting method called Crafting the Path tailored for retrieval systems. Crafting the Path involves a three-step process that crafts query-related information necessary for finding the passages to be searched in each step. Specifically, the Crafting the Path begins with Query Concept Comprehension, proceeds to Query Type Identification, and finally conducts Expected Answer Extraction. Experimental results show that our method outperforms previous rewriting methods, especially in less familiar domains for LLMs. We demonstrate that our method is less dependent on the internal parameter knowledge of the model and generates queries with fewer factual inaccuracies. Furthermore, we observe that Crafting the Path has less latency compared to the baselines.

The leak was first reported in the gaming press and then picked up by the Wall Street Journal, which said some of the leaked material related to advertising campaigns and interview candidates, with some dating back as far as 2019. There has been growing concern amongst performers, artists and other creatives that the rapid spread of generative AI will undermine their livelihoods and damage the creative environment. Generative AI is trained on vast bodies of existing material - including texts, images, music and video. It is then able to produce new work of a standard that can be hard to distinguish from human-generated material. Nullbulge describes itself as "a hacktivist group protecting artists' rights and ensuring fair compensation for their work".

On today's episode of Hear Me Out: tried and Turing tested. Coming into the 2024 election cycle, generative A.I. was one of the main concerns for democracy watchdogs; its power to create deceptive text, images, and sounds at a rapid, unfettered pace seems ripe to spread misinformation. But of all the controversies and current events that have shaped the election thus far… A.I., somehow, might not be one of them. Writer and social strategist Rachel Greenspan joins us to share what she's hearing about the A.I. revolution that wasn't. If you have thoughts you want to share, or an idea for a topic we should tackle, you can email the show: hearmeout@slate.com

WIRED's coverage of the best Amazon Prime Day deals is, as they say, built different. For starters, we only include products someone from our team has personally tested and reviewed. That means you will not find any flimsy fad gadgets or shoddy dupes among our recommendations. What remains is all solid stuff. You'll often find a link to a longer write-up to a review or buying guide if you want to make a fully informed buying decision. Additionally, we obsessively track prices to make sure everything on the list is a genuinely good price right now. For more on that, consult our helpful guide to shopping like a pro on Prime Day. We test products year-round and handpicked these Prime Day deals. Products that are sold out or no longer discounted will be crossed out. We'll update this guide regularly throughout Prime Day by adding fresh deals and removing dead deals. Logitech makes a lot of great, functional keyboards, but the Pop Keys (9/10, WIRED Recommends) not only leverage the ...

Robust access to trustworthy information is a critical need for society including implications for knowledge production, public health education, and promoting informed citizenry in democratic societies. Generative AI technologies such as large language models (LLMs) may enable new ways to access information and improve effectiveness of existing information retrieval (IR) systems. More efficient basic task execution with the help of LLMs can also enable people to focus on the more challenging aspects of information retrieval related tasks and research. However, the long-term social implications of deploying these technologies in the context of information access are not yet well-understood. Existing research has focused on how these models may generate biased and harmful content [11, 23, 69, 80, 124, 158, 236] as well as the environmental costs [23, 31, 61, 166, 167, 241] of developing and deploying these models at scale. In the context of information access, Shah and Bender [187] have argued that certain framings of LLMs as "search engines" lack the necessary theoretical underpinnings and may constitute as a category error. In this current work, we present a broader perspective on the sociotechnical implications of generative AI for information access. Our perspective is informed by existing literature and aims to provide a summary of known challenges viewed through a systemic lens that we hope will serve as a useful resource for future critical research in this area. We present a summary of these implications next followed by recommendations for evaluation and mitigation later in this chapter.

The Audio-Visual Question Answering (AVQA) task holds significant potential for applications. Compared to traditional unimodal approaches, the multi-modal input of AVQA makes feature extraction and fusion processes more challenging. Euclidean space is difficult to effectively represent multi-dimensional relationships of data. Especially when extracting and processing data with a tree structure or hierarchical structure, Euclidean space is not suitable as an embedding space. Additionally, the self-attention mechanism in Transformers is effective in capturing the dynamic relationships between elements in a sequence. However, the self-attention mechanism's limitations in window modeling and quadratic computational complexity reduce its effectiveness in modeling long sequences. To address these limitations, we propose SHMamba: Structured Hyperbolic State Space Model to integrate the advantages of hyperbolic geometry and state space models. Specifically, SHMamba leverages the intrinsic properties of hyperbolic space to represent hierarchical structures and complex relationships in audio-visual data. Meanwhile, the state space model captures dynamic changes over time by globally modeling the entire sequence. Furthermore, we introduce an adaptive curvature hyperbolic alignment module and a cross fusion block to enhance the understanding of hierarchical structures and the dynamic exchange of cross-modal information, respectively. Extensive experiments demonstrate that SHMamba outperforms previous methods with fewer parameters and computational costs. Our learnable parameters are reduced by 78.12\%, while the average performance improves by 2.53\%. Experiments show that our method demonstrates superiority among all current major methods and is more suitable for practical application scenarios.

For web agents to be practically useful, they must adapt to the continuously evolving web environment characterized by frequent updates to user interfaces and content. However, most existing benchmarks only capture the static aspects of the web. To bridge this gap, we introduce WebCanvas, an innovative online evaluation framework for web agents that effectively addresses the dynamic nature of web interactions. WebCanvas contains three main components to facilitate realistic assessments: (1) A novel evaluation metric which reliably capture critical intermediate actions or states necessary for task completions while disregarding noise caused by insignificant events or changed web-elements. (2) A benchmark dataset called Mind2Web-Live, a refined version of original Mind2Web static dataset containing 542 tasks with 2439 intermediate evaluation states; (3) Lightweight and generalizable annotation tools and testing pipelines that enables the community to collect and maintain the high-quality, up-to-date dataset. Building on WebCanvas, we open-source an agent framework with extensible modules for reasoning, providing a foundation for the community to conduct online inference and evaluations. Our best-performing agent achieves a task success rate of 23.1% and a task completion rate of 48.8% on the Mind2Web-Live test set. Additionally, we analyze the performance discrepancies across various websites, domains, and experimental environments. We encourage the community to contribute further insights on online agent evaluation, thereby advancing this field of research.

Large visual-language models (LVLMs) exhibit exceptional performance in visual-language reasoning across diverse cross-modal benchmarks. Despite these advances, recent research indicates that Large Language Models (LLMs), like GPT-3.5-turbo, underachieve compared to well-trained smaller models, such as BERT, in Fake News Detection (FND), prompting inquiries into LVLMs' efficacy in FND tasks. Although performance could improve through fine-tuning LVLMs, the substantial parameters and requisite pre-trained weights render it a resource-heavy endeavor for FND applications. This paper initially assesses the FND capabilities of two notable LVLMs, CogVLM and GPT4V, in comparison to a smaller yet adeptly trained CLIP model in a zero-shot context. The findings demonstrate that LVLMs can attain performance competitive with that of the smaller model. Next, we integrate standard in-context learning (ICL) with LVLMs, noting improvements in FND performance, though limited in scope and consistency. To address this, we introduce the \textbf{I}n-context \textbf{M}ultimodal \textbf{F}ake \textbf{N}ews \textbf{D}etection (IMFND) framework, enriching in-context examples and test inputs with predictions and corresponding probabilities from a well-trained smaller model. This strategic integration directs the LVLMs' focus towards news segments associated with higher probabilities, thereby improving their analytical accuracy. The experimental results suggest that the IMFND framework significantly boosts the FND efficiency of LVLMs, achieving enhanced accuracy over the standard ICL approach across three publicly available FND datasets.