The unique capabilities of Large Language Models (LLMs), such as the natural language text generation ability, position them as strong candidates for providing explanation for recommendations. However, despite the size of the LLM, most existing models struggle to produce zero-shot explanations reliably. To address this issue, we propose a framework called Logic-Scaffolding, that combines the ideas of aspect-based explanation and chain-of-thought prompting to generate explanations through intermediate reasoning steps. In this paper, we share our experience in building the framework and present an interactive demonstration for exploring our results.

Large language models (LLMs) have exploded in popularity due to their ability to perform a wide array of natural language tasks. Text-based content moderation is one LLM use case that has received recent enthusiasm, however, there is little research investigating how LLMs perform in content moderation settings. In this work, we evaluate a suite of commodity LLMs on two common content moderation tasks: rule-based community moderation and toxic content detection. For rule-based community moderation, we instantiate 95 subcommunity specific LLMs by prompting GPT-3.5 with rules from 95 Reddit subcommunities. We find that GPT-3.5 is effective at rule-based moderation for many communities, achieving a median accuracy of 64% and a median precision of 83%. For toxicity detection, we evaluate a suite of commodity LLMs (GPT-3, GPT-3.5, GPT-4, Gemini Pro, LLAMA 2) and show that LLMs significantly outperform currently widespread toxicity classifiers. However, recent increases in model size add only marginal benefit to toxicity detection, suggesting a potential performance plateau for LLMs on toxicity detection tasks. We conclude by outlining avenues for future work in studying LLMs and content moderation.

As large language models (LLMs) like ChatGPT have gained traction, an increasing number of news websites have begun utilizing them to generate articles. However, not only can these language models produce factually inaccurate articles on reputable websites but disreputable news sites can utilize LLMs to mass produce misinformation. To begin to understand this phenomenon, we present one of the first large-scale studies of the prevalence of synthetic articles within online news media. To do this, we train a DeBERTa-based synthetic news detector and classify over 15.90 million articles from 3,074 misinformation and mainstream news websites. We find that between January 1, 2022, and May 1, 2023, the relative number of synthetic news articles increased by 55.4% on mainstream websites while increasing by 457% on misinformation sites. We find that this increase is largely driven by smaller less popular websites. Analyzing the impact of the release of ChatGPT using an interrupted-time-series, we show that while its release resulted in a marked increase in synthetic articles on small sites as well as misinformation news websites, there was not a corresponding increase on large mainstream news websites.

With pre-orders for the Apple Vision Pro headset opening this week, the company is getting ready to launch one of its most significant products ever. It announced this morning an "entertainment format pioneered by Apple" called Apple Immersive Video, as well as new viewing environments in the Disney app featuring scenes from the studio's beloved franchises like the Avengers and Star Wars. We already got hands-on once back at WWDC when the headset was first announced, but two of our editors, Dana Wollman and Cherlynn Low, had a chance to go back and revisit the device (and Dana's case, experience it anew). Since we've already walked you through some of the basic UI elements in our earlier piece, we decided to focus on some of the more recently added features, including Apple Immersive Video, the new Disney environments, a built-in "Encounter Dinosaurs" experience, as well as the floating keyboard, which didn't work for us when we first tried the device in June of last year. Here, too, we wanted to really get at what it actually feels like to use the device, from the frustrating to the joyful to the unintentionally eerie.

An academic author and former Harvard professor defended former Harvard President Claudine Gay after her plagiarism scandal, arguing citation mistakes are commonly found in academic work. "Essentially, what I have to say is what a lot of people have said. What she did was a minor infraction of the rules," Dr. Marshall Poe told Fox News Digital. Poe spent years teaching Russian and Eurasian history at elite universities like Harvard and has published interviews with thousands of scholars for his podcast platform, the New Books Network. An academic author himself, Poe argues true "idea theft" in academia is exceedingly rare, and he doesn't believe Gay was guilty of it.

As the US gears up for the 2024 presidential election, OpenAI shares its plans on suppressing misinformation related to elections worldwide, with a focus set on boosting the transparency around the origin of information. One such highlight is the use of cryptography -- as standardized by the Coalition for Content Provenance and Authenticity -- to encode the provenance of images generated by DALL-E 3. This will allow the platform to better detect AI-generated images using a provenance classifier, in order to help voters assess the reliability of certain content. This approach is similar to, if not better than, DeepMind's SynthID for digitally watermark AI-generated images and audio, as part of Google's own election content strategy published last month. Meta's AI image generator also adds an invisible watermark to its content, though the company has yet to share its readiness on tackling election-related misinformation.

OpenAI is rolling out a series of initiatives to prevent its products from being used for misinformation ahead of a major year for elections globally. On Monday, the artificial intelligence startup announced new tools that will attribute information about current events provided by its chatbot ChatGPT, and help users determine if an image was created by its AI software. The changes come as concerns rise over the risks of so-called deepfakes -- manipulated videos or other digital representations -- and other AI-produced content that could misguide voters during campaigns. "Protecting the integrity of elections requires collaboration from every corner of the democratic process, and we want to make sure our technology is not used in a way that could undermine this process," the company wrote in a blog post on Monday.

Audio embeddings enable large scale comparisons of the similarity of audio files for applications such as search and recommendation. Due to the subjectivity of audio similarity, it can be desirable to design systems that answer not only whether audio is similar, but similar in what way (e.g., wrt. tempo, mood or genre). Previous works have proposed disentangled embedding spaces where subspaces representing specific, yet possibly correlated, attributes can be weighted to emphasize those attributes in downstream tasks. However, no research has been conducted into the independence of these subspaces, nor their manipulation, in order to retrieve tracks that are similar but different in a specific way. Here, we explore the manipulation of tempo in embedding spaces as a case-study towards this goal. We propose tempo translation functions that allow for efficient manipulation of tempo within a pre-existing embedding space whilst maintaining other properties such as genre. As this translation is specific to tempo it enables retrieval of tracks that are similar but have specifically different tempi. We show that such a function can be used as an efficient data augmentation strategy for both training of downstream tempo predictors, and improved nearest neighbor retrieval of properties largely independent of tempo.

This paper addresses the problem of global tempo estimation in musical audio. Given that annotating tempo is time-consuming and requires certain musical expertise, few publicly available data sources exist to train machine learning models for this task. Towards alleviating this issue, we propose a fully self-supervised approach that does not rely on any human labeled data. Our method builds on the fact that generic (music) audio embeddings already encode a variety of properties, including information about tempo, making them easily adaptable for downstream tasks. While recent work in self-supervised tempo estimation aimed to learn a tempo specific representation that was subsequently used to train a supervised classifier, we reformulate the task into the binary classification problem of predicting whether a target track has the same or a different tempo compared to a reference. While the former still requires labeled training data for the final classification model, our approach uses arbitrary unlabeled music data in combination with time-stretching for model training as well as a small set of synthetically created reference samples for predicting the final tempo. Evaluation of our approach in comparison with the state-of-the-art reveals highly competitive performance when the constraint of finding the precise tempo octave is relaxed.

Audio embeddings are crucial tools in understanding large catalogs of music. Typically embeddings are evaluated on the basis of the performance they provide in a wide range of downstream tasks, however few studies have investigated the local properties of the embedding spaces themselves which are important in nearest neighbor algorithms, commonly used in music search and recommendation. In this work we show that when learning audio representations on music datasets via contrastive learning, musical properties that are typically homogeneous within a track (e.g., key and tempo) are reflected in the locality of neighborhoods in the resulting embedding space. By applying appropriate data augmentation strategies, localisation of such properties can not only be reduced but the localisation of other attributes is increased. For example, locality of features such as pitch and tempo that are less relevant to non-expert listeners, may be mitigated while improving the locality of more salient features such as genre and mood, achieving state-of-the-art performance in nearest neighbor retrieval accuracy. Similarly, we show that the optimal selection of data augmentation strategies for contrastive learning of music audio embeddings is dependent on the downstream task, highlighting this as an important embedding design decision.