Whether you dream of holstering a flintlock pistol and sailing through the 18th-century Golden Age of Piracy or leading a clan of Vikings to settle in the fractured Anglo-Saxon kingdoms of the 9th century, Assassin's Creed video games have you covered. Since 2007, the popular action-adventure series created by video game publisher Ubisoft has been taking gamers on adventures around the globe through different historical periods. With its 13th instalment released on Thursday, Assassin's Creed Mirage attempts to immerse players in Iraq's 9th-century Baghdad during the rule of the Abbasid Caliphate, when it was one of the most significant cities in the world. Today's capital of Iraq is often associated, especially by those in the West, with the United States war and the destruction it brought more than two decades ago. But in Assassin's Creed Mirage, the game attempts to give players a glimpse into the rich and diverse history of the Abbasid Caliphate during the Islamic Golden Age.

Text-to-image model personalization aims to introduce a user-provided concept to the model, allowing its synthesis in diverse contexts. However, current methods primarily focus on the case of learning a single concept from multiple images with variations in backgrounds and poses, and struggle when adapted to a different scenario. In this work, we introduce the task of textual scene decomposition: given a single image of a scene that may contain several concepts, we aim to extract a distinct text token for each concept, enabling fine-grained control over the generated scenes. To this end, we propose augmenting the input image with masks that indicate the presence of target concepts. These masks can be provided by the user or generated automatically by a pre-trained segmentation model. We then present a novel two-phase customization process that optimizes a set of dedicated textual embeddings (handles), as well as the model weights, striking a delicate balance between accurately capturing the concepts and avoiding overfitting. We employ a masked diffusion loss to enable handles to generate their assigned concepts, complemented by a novel loss on cross-attention maps to prevent entanglement. We also introduce union-sampling, a training strategy aimed to improve the ability of combining multiple concepts in generated images. We use several automatic metrics to quantitatively compare our method against several baselines, and further affirm the results using a user study. Finally, we showcase several applications of our method. Project page is available at: https://omriavrahami.com/break-a-scene/

Contrastive Language-Image Pre-training (CLIP) is an approach that has advanced research and applications in computer vision, fueling modern recognition systems and generative models. We believe that the main ingredient to the success of CLIP is its data and not the model architecture or pre-training objective. However, CLIP only provides very limited information about its data and how it has been collected, leading to works that aim to reproduce CLIP's data by filtering with its model parameters. In this work, we intend to reveal CLIP's data curation approach and in our pursuit of making it open to the community introduce Metadata-Curated Language-Image Pre-training (MetaCLIP). MetaCLIP takes a raw data pool and metadata (derived from CLIP's concepts) and yields a balanced subset over the metadata distribution. Our experimental study rigorously isolates the model and training settings, concentrating solely on data. MetaCLIP applied to CommonCrawl with 400M image-text data pairs outperforms CLIP's data on multiple standard benchmarks. In zero-shot ImageNet classification, MetaCLIP achieves 70.8% accuracy, surpassing CLIP's 68.3% on ViT-B models. Scaling to 1B data, while maintaining the same training budget, attains 72.4%. Our observations hold across various model sizes, exemplified by ViT-H achieving 80.5%, without any bells-and-whistles. Curation code and training data distribution on metadata is made available at https://github.com/facebookresearch/MetaCLIP.

The field of urban spatial-temporal prediction is advancing rapidly with the development of deep learning techniques and the availability of large-scale datasets. However, challenges persist in accessing and utilizing diverse urban spatial-temporal datasets from different sources and stored in different formats, as well as determining effective model structures and components with the proliferation of deep learning models. This work addresses these challenges and provides three significant contributions. Firstly, we introduce "atomic files", a unified storage format designed for urban spatial-temporal big data, and validate its effectiveness on 40 diverse datasets, simplifying data management. Secondly, we present a comprehensive overview of technological advances in urban spatial-temporal prediction models, guiding the development of robust models. Thirdly, we conduct extensive experiments using diverse models and datasets, establishing a performance leaderboard and identifying promising research directions. Overall, this work effectively manages urban spatial-temporal data, guides future efforts, and facilitates the development of accurate and efficient urban spatial-temporal prediction models. It can potentially make long-term contributions to urban spatial-temporal data management and prediction, ultimately leading to improved urban living standards.

Time-series causal discovery (TSCD) is a fundamental problem of machine learning. However, existing synthetic datasets cannot properly evaluate or predict the algorithms' performance on real data. This study introduces the CausalTime pipeline to generate time-series that highly resemble the real data and with ground truth causal graphs for quantitative performance evaluation. The pipeline starts from real observations in a specific scenario and produces a matching benchmark dataset. Firstly, we harness deep neural networks along with normalizing flow to accurately capture realistic dynamics. Secondly, we extract hypothesized causal graphs by performing importance analysis on the neural network or leveraging prior knowledge. Thirdly, we derive the ground truth causal graphs by splitting the causal model into causal term, residual term, and noise term. Lastly, using the fitted network and the derived causal graph, we generate corresponding versatile time-series proper for algorithm assessment. In the experiments, we validate the fidelity of the generated data through qualitative and quantitative experiments, followed by a benchmarking of existing TSCD algorithms using these generated datasets. CausalTime offers a feasible solution to evaluating TSCD algorithms in real applications and can be generalized to a wide range of fields. For easy use of the proposed approach, we also provide a user-friendly website, hosted on www.causaltime.cc.

As deep learning technology advances and more urban spatial-temporal data accumulates, an increasing number of deep learning models are being proposed to solve urban spatial-temporal prediction problems. However, there are limitations in the existing field, including open-source data being in various formats and difficult to use, few papers making their code and data openly available, and open-source models often using different frameworks and platforms, making comparisons challenging. A standardized framework is urgently needed to implement and evaluate these methods. To address these issues, we propose LibCity, an open-source library that offers researchers a credible experimental tool and a convenient development framework. In this library, we have reproduced 65 spatial-temporal prediction models and collected 55 spatial-temporal datasets, allowing researchers to conduct comprehensive experiments conveniently. By enabling fair model comparisons, designing a unified data storage format, and simplifying the process of developing new models, LibCity is poised to make significant contributions to the spatial-temporal prediction field.

Recent developments in large language models (LLMs) have been impressive. However, these models sometimes show inconsistencies and problematic behavior, such as hallucinating facts, generating flawed code, or creating offensive and toxic content. Unlike these models, humans typically utilize external tools to cross-check and refine their initial content, like using a search engine for fact-checking, or a code interpreter for debugging. Inspired by this observation, we introduce a framework called CRITIC that allows LLMs, which are essentially "black boxes" to validate and progressively amend their own outputs in a manner similar to human interaction with tools. More specifically, starting with an initial output, CRITIC interacts with appropriate tools to evaluate certain aspects of the text, and then revises the output based on the feedback obtained during this validation process. Comprehensive evaluations involving free-form question answering, mathematical program synthesis, and toxicity reduction demonstrate that CRITIC consistently enhances the performance of LLMs. Meanwhile, our research highlights the crucial importance of external feedback in promoting the ongoing self-improvement of LLMs.

As Republican presidential candidates flocked to Southern California for a debate and the state GOP convention this week, President Biden was busy in the San Francisco Bay Area collecting campaign checks and painting the election as a choice between MAGA chaos and functioning government. During three fundraising events in some of the swankiest neighborhoods of Silicon Valley and San Francisco on Tuesday and Wednesday, Biden touted his administration's accomplishments on climate and infrastructure, the United States' support for Ukraine and opposition to Russian President Vladimir Putin and his appointment of the first Black woman to the Supreme Court. As Wednesday's GOP presidential debate at the Reagan library made clear, the Republican Party has moved a long way from Reaganism. Despite that progress for the Democratic Party, Biden said he was running for reelection because "democracy is still at stake" in next year's election, a likely rematch with former President Trump, who has surged ahead in Republican primary polls. "Donald Trump and the MAGA Republicans are determined to destroy this democracy," Biden said during a private fundraising event Tuesday evening at the Atherton mansion of Democratic donors and philanthropists Liz Simons and Mark Heising in San Francisco.

The increasing capabilities of large language models (LLMs) raise opportunities for artificial general intelligence but concurrently amplify safety concerns, such as potential misuse of AI systems, necessitating effective AI alignment. Reinforcement Learning from Human Feedback (RLHF) has emerged as a promising pathway towards AI alignment but brings forth challenges due to its complexity and dependence on a separate reward model. Direct Preference Optimization (DPO) has been proposed as an alternative, and it remains equivalent to RLHF under the reverse KL regularization constraint. This paper presents $f$-DPO, a generalized approach to DPO by incorporating diverse divergence constraints. We show that under certain $f$-divergences, including Jensen-Shannon divergence, forward KL divergences and $\alpha$-divergences, the complex relationship between the reward and optimal policy can also be simplified by addressing the Karush-Kuhn-Tucker conditions. This eliminates the need for estimating the normalizing constant in the Bradley-Terry model and enables a tractable mapping between the reward function and the optimal policy. Our approach optimizes LLMs to align with human preferences in a more efficient and supervised manner under a broad set of divergence constraints. Empirically, adopting these divergences ensures a balance between alignment performance and generation diversity. Importantly, $f$-DPO outperforms PPO-based methods in divergence efficiency, and divergence constraints directly influence expected calibration error (ECE).

The possibility of automatically classifying high frequency sub-bottom acoustic reflections collected from an Autonomous Underwater Robot is investigated in this paper. In field surveys of Cobalt-rich Manganese Crusts (Mn-crusts), existing methods relies on visual confirmation of seafloor from images and thickness measurements using the sub-bottom probe. Using these visual classification results as ground truth, an autoencoder is trained to extract latent features from bundled acoustic reflections. A Support Vector Machine classifier is then trained to classify the latent space to idetify seafloor classes. Results from data collected from seafloor at 1500m deep regions of Mn-crust showed an accuracy of about 70%.