We are living in a three-dimensional space while moving forward through a fourth dimension: time. To allow artificial intelligence to develop a comprehensive understanding of such a 4D environment, we introduce 4D Panoptic Scene Graph (PSG-4D), a new representation that bridges the raw visual data perceived in a dynamic 4D world and high-level visual understanding. Specifically, PSG-4D abstracts rich 4D sensory data into nodes, which represent entities with precise location and status information, and edges, which capture the temporal relations. To facilitate research in this new area, we build a richly annotated PSG-4D dataset consisting of 3K RGB-D videos with a total of 1M frames, each of which is labeled with 4D panoptic segmentation masks as well as fine-grained, dynamic scene graphs. To solve PSG-4D, we propose PSG4DFormer, a Transformer-based model that can predict panoptic segmentation masks, track masks along the time axis, and generate the corresponding scene graphs via a relation component. Extensive experiments on the new dataset show that our method can serve as a strong baseline for future research on PSG-4D.

Elon Musk has launched a 97.4 billion bid for AI darling OpenAI. The Wall Street Journal reported that a group of investors led by Musk's xAI submitted an unsolicited offer to the company's board of directors on Monday. It's a bid for the non-profit that controls OpenAI's for-profit arm. OpenAI is not a traditional company, and the non-profit structure Sam Altman and others at the company want it to get away from may, in fact, protect it from Musk's offer. There's further drama around all this: Musk had sued OpenAI and Sam Altman for allegedly ditching its non-profit mission around this time last year.

Grindr's AI wingman, currently in beta testing with around 10,000 users, arrives at a pivotal moment for the software company. With its iconic notification chirp and ominous mask logo, the app is known culturally as a digital bathhouse for gay and bisexual men to swap nudes and meet with nearby users for sex, but Grindr CEO George Arison sees the addition of a generative AI assistant and machine intelligence tools as an opportunity for expansion. "This is not just a hookup product anymore," he says. "There's obviously no question that it started out as a hookup product, but the fact that it's become a lot more over time is something people don't fully appreciate." Grindr's product road map for 2025 spotlights multiple AI features aimed at current power users, like chat summaries, as well as dating and travel-focused tools.

The widespread adoption of Transformer architectures in various data modalities has opened new avenues for the applications in molecular modeling. Nevertheless, it remains elusive that whether the Transformer-based architecture can do molecular modeling as good as equivariant GNNs. In this paper, by designing Interatomic Positional Encoding (IPE) thatparameterizes atomic environments as Transformer's positional encodings,we propose Geoformer, a novel geometric Transformer to effectively model molecular structures for various molecular property prediction. We evaluate Geoformer on several benchmarks, including the QM9 dataset and the recently proposed Molecule3D dataset.

Similar to natural language models, researchers have proposed foundation models in genomics to learn generalizable features from unlabeled genome data that can then be fine-tuned for downstream tasks such as identifying regulatory elements. Due to the quadratic scaling of attention, previous Transformer-based genomic models have used 512 to 4k tokens as context ( 0.001% of the human genome), significantly limiting the modeling of long-range interactions in DNA. In addition, these methods rely on tokenizers or fixed k-mers to aggregate meaningful DNA units, losing single nucleotide resolution (i.e. DNA "characters") where subtle genetic variations can completely alter protein function via single nucleotide polymorphisms (SNPs). Recently, Hyena, a large language model based on implicit convolutions was shown to match attention in quality while allowing longer context lengths and lower time complexity.

A group of investors led by Elon Musk is offering about 97.4 billion to buy the nonprofit behind OpenAI, escalating a dispute with the artificial intelligence company that Musk helped found a decade ago. Musk and his own AI startup, xAI, and a consortium of investment firms want to take control of the ChatGPT maker and revert it to its original charitable mission as a nonprofit research lab, according to Musk's attorney Marc Toberoff. OpenAI CEO Sam Altman quickly rejected the unsolicited bid on Musk's social platform X, saying, "no thank you but we will buy Twitter for 9.74 billion if you want." Musk bought Twitter, now called X, for 44 billion in 2022. Musk and Altman, who together helped start OpenAI in 2015 and later competed over who should lead it, have been in a long-running feud over the startup's direction since Musk resigned from its board in 2018.

A consortium led by Elon Musk said on Monday it has offered 97.4 billion to buy the nonprofit that controls OpenAI, another salvo in the billionaire's fight to block the artificial intelligence startup from transitioning to a for-profit firm. Musk's bid is likely to ratchet up longstanding tensions with OpenAI CEO Sam Altman over the future of the ChatGPT maker at the heart of a boom in generative AI technology. Altman on Monday promptly posted on X: "no thank you but we will buy twitter for 9.74 billion if you want." Musk cofounded OpenAI with Altman in 2015 as a nonprofit, but left before the company took off. He founded the competing AI startup xAI in 2023.

Large language models (LLMs) have achieved significant performance gains via scaling up model sizes and/or data. However, recent evidence suggests diminishing returns from such approaches, motivating scaling the computation spent at inference time. Existing inference-time scaling methods, usually with reward models, cast the task as a search problem, which tends to be vulnerable to reward hacking as a consequence of approximation errors in reward models. In this paper, we instead cast inference-time scaling as a probabilistic inference task and leverage sampling-based techniques to explore the typical set of the state distribution of a state-space model with an approximate likelihood, rather than optimize for its mode directly. We propose a novel inference-time scaling approach by adapting particle-based Monte Carlo methods to this task. Our empirical evaluation demonstrates that our methods have a 4-16x better scaling rate over our deterministic search counterparts on various challenging mathematical reasoning tasks. Using our approach, we show that Qwen2.5-Math-1.5B-Instruct can surpass GPT-4o accuracy in only 4 rollouts, while Qwen2.5-Math-7B-Instruct scales to o1 level accuracy in only 32 rollouts. Our work not only presents an effective method to inference-time scaling, but also connects the rich literature in probabilistic inference with inference-time scaling of LLMs to develop more robust algorithms in future work. Code, videos, and further information available at https://probabilistic-inference-scaling.github.io.

Today's gen-AI workflows that involve multiple ML model calls, tool/API calls, data retrieval, or generic code execution are often tuned manually in an ad-hoc way that is both time-consuming and error-prone. In this paper, we propose a systematic approach for automatically tuning gen-AI workflows. Our key insight is that gen-AI workflows can benefit from structure, operator, and prompt changes, but unique properties of gen-AI workflows require new optimization techniques. We propose AdaSeek, an adaptive hierarchical search algorithm for autotuning gen-AI workflows. AdaSeek organizes workflow tuning methods into different layers based on the user-specified total search budget and distributes the budget across different layers based on the complexity of each layer. During its hierarchical search, AdaSeek redistributes the search budget from less useful to more promising tuning configurations based on workflow-level evaluation results. We implement AdaSeek in a workflow autotuning framework called Cognify and evaluate Cognify using six types of workflows such as RAG-based QA and text-to-SQL transformation. Overall, Cognify improves these workflows' generation quality by up to 2.8x, reduces execution monetary cost by up to 10x, and reduces end-to-end latency by 2.7x.

Growing concerns about climate change and sustainability are driving manufacturers to take significant steps toward reducing their carbon footprints. For these manufacturers, a first step towards this goal is to identify the environmental impact of the individual components of their products. We propose a system leveraging large language models (LLMs) to automatically map components from manufacturer Bills of Materials (BOMs) to Life Cycle Assessment (LCA) database entries by using LLMs to expand on available component information. Our approach reduces the need for manual data processing, paving the way for more accessible sustainability practices.