The advancement of LLMs has significantly boosted the performance of complex long-form question answering tasks. However, one prominent issue of LLMs is the generated "hallucination" responses that are not factual. Consequently, attribution for each claim in responses becomes a common solution to improve the factuality and verifiability. Existing researches mainly focus on how to provide accurate citations for the response, which largely overlook the importance of identifying the claims or statements for each response. To bridge this gap, we introduce a new claim decomposition benchmark, which requires building system that can identify atomic and checkworthy claims for LLM responses. Specifically, we present the Chinese Atomic Claim Decomposition Dataset (CACDD), which builds on the WebCPM dataset with additional expert annotations to ensure high data quality. The CACDD encompasses a collection of 500 human-annotated question-answer pairs, including a total of 4956 atomic claims. We further propose a new pipeline for human annotation and describe the challenges of this task. In addition, we provide experiment results on zero-shot, few-shot and fine-tuned LLMs as baselines. The results show that the claim decomposition is highly challenging and requires further explorations. All code and data are publicly available at \url{https://github.com/FBzzh/CACDD}.

Partial-differential-equation (PDE)-constrained optimization is a well-worn technique for acquiring optimal parameters of systems governed by PDEs. However, this approach is limited to providing a single set of optimal parameters per optimization. Given a differentiable PDE solver, if the free parameters are reparameterized as the output of a neural network, that neural network can be trained to learn a map from a probability distribution to the distribution of optimal parameters. This proves useful in the case where there are many well performing local minima for the PDE. We apply this technique to train a neural network that generates optimal parameters that minimize laser-plasma instabilities relevant to laser fusion and show that the neural network generates many well performing and diverse minima.

The widespread deployment of sensing devices leads to a surge in data for spatio-temporal forecasting applications such as traffic flow, air quality, and wind energy. Although spatio-temporal graph neural networks have achieved success in modeling various static spatio-temporal forecasting scenarios, real-world spatio-temporal data are typically received in a streaming manner, and the network continuously expands with the installation of new sensors. Thus, spatio-temporal forecasting in streaming scenarios faces dual challenges: the inefficiency of retraining models over newly arrived data and the detrimental effects of catastrophic forgetting over long-term history. To address these challenges, we propose a novel prompt tuning-based continuous forecasting method, following two fundamental tuning principles guided by empirical and theoretical analysis: expand and compress, which effectively resolve the aforementioned problems with lightweight tuning parameters. Specifically, we integrate the base spatio-temporal graph neural network with a continuous prompt pool, utilizing stored prompts (i.e., few learnable parameters) in memory, and jointly optimize them with the base spatio-temporal graph neural network. This method ensures that the model sequentially learns from the spatio-temporal data stream to accomplish tasks for corresponding periods. Extensive experimental results on multiple real-world datasets demonstrate the multi-faceted superiority of our method over the state-of-the-art baselines, including effectiveness, efficiency, universality, etc.

Generating realistic time series samples is crucial for stress-testing models and protecting user privacy by using synthetic data. In engineering and safety-critical applications, these samples must meet certain hard constraints that are domainspecific or naturally imposed by physics or nature. Consider, for example, generating electricity demand patterns with constraints on peak demand times. This can be used to stress-test the functioning of power grids during adverse weather conditions. Existing approaches for generating constrained time series are either not scalable or degrade sample quality. To address these challenges, we introduce Constrained Posterior Sampling (CPS), a diffusion-based sampling algorithm that aims to project the posterior mean estimate into the constraint set after each denoising update. We provide theoretical justifications highlighting the impact of our projection step on sampling. Empirically, CPS outperforms state-of-the-art methods in sample quality and similarity to real time series by around 10% and 42%, respectively, on real-world stocks, traffic, and air quality datasets. Synthesizing realistic time series samples can aid in "what-if" scenario analysis, stress-testing machine learning (ML) models (Rizzato et al., 2022; Gowal et al., 2021), anonymizing private user data (Yoon et al., 2020), etc. Current approaches for time series generation use state-of-the-art (SOTA) generative models, such as Generative Adversarial Networks (GANs) (Yoon et al., 2019; Donahue et al., 2018) and Diffusion Models (DMs) (Tashiro et al., 2021; Alcaraz & Strodthoff, 2023; Narasimhan et al., 2024), to generate high-fidelity time series samples. GPT-4 (Bubeck et al., 2023) and Stable Diffusion (Podell et al., 2023), has increased the focus on constraining the outputs from these models, Note that we cannot clearly define the notion of a constraint set in these domains. For example, verifying if the image of a hand has 6 fingers is practically hard, as all deep-learned perception models for this task have associated prediction errors. However, our key insight is that we can describe a time series through statistical features computed using well-defined functions.

Periodic soft real-time systems have broad applications in many areas, such as IoT. Finding an optimal energy-efficient policy that is adaptable to underlying edge devices while meeting deadlines for tasks has always been challenging. This research studies generalized systems with multi-task, multi-deadline scenarios with reinforcement learning-based DVFS for energy saving. This work addresses the limitation of previous work that models a periodic system as a single task and single-deadline scenario, which is too simplified to cope with complex situations. The method encodes time series information in the Linux kernel into information that is easy to use for reinforcement learning, allowing the system to generate DVFS policies to adapt system patterns based on the general workload. For encoding, we present two different methods for comparison. Both methods use only one performance counter: system utilization and the kernel only needs minimal information from the userspace. Our method is implemented on Jetson Nano Board (2GB) and is tested with three fixed multitask workloads, which are three, five, and eight tasks in the workload, respectively. For randomness and generalization, we also designed a random workload generator to build different multitask workloads to test. Based on the test results, our method could save 3%-10% power compared to Linux built-in governors.

Recent normalization-based methods have shown great success in tackling the distribution shift issue, facilitating non-stationary time series forecasting. Since these methods operate in the time domain, they may fail to fully capture the dynamic patterns that are more apparent in the frequency domain, leading to suboptimal results. This paper first theoretically analyzes how normalization methods affect frequency components. We prove that the current normalization methods that operate in the time domain uniformly scale non-zero frequencies, and thus, they struggle to determine components that contribute to more robust forecasting. Therefore, we propose FredNormer, which observes datasets from a frequency perspective and adaptively up-weights the key frequency components. To this end, FredNormer consists of two components: a statistical metric that normalizes the input samples based on their frequency stability and a learnable weighting layer that adjusts stability and introduces sample-specific variations. Notably, FredNormer is a plug-and-play module, which does not compromise the efficiency compared to existing normalization methods. Extensive experiments show that FredNormer improves the averaged MSE of backbone forecasting models by 33.3% and 55.3% on the ETTm2 dataset. Compared to the baseline normalization methods, FredNormer achieves 18 top-1 results and 6 top-2 results out of 28 settings.

Google is officially putting its weight behind advanced "mini" nuclear reactors in an effort to produce new clean to meet growing AI energy demands. On Tuesday, the company announced an agreement with California-based small nuclear reactor (SMR) startup Kairos Power to commission the development of six or seven reactors that could add 500 megawatts of clean energy to the US electrical grid within the next decade. Google's buy-in represents the biggest investment for the experimental new reactor type from a tech company and could play a key in making so-called next-generation nuclear commercially viable. The deal is part of a broader embrace of nuclear power by tech giants who are frantically searching for ways to fuel their increasing energy consumption while attempting to stick to their climate goals. In a blog post, Google said it expects the first of Kairos reactors to come online as early as 2030, with the other five six operational by 2035.

With its Gemini chatbot and Pixel AI phone software, it's fair to say Google has an obsessive focus on artificial intelligence. But all that advanced computational power requires millions of computers, known as'servers', housed inside data centres across the world that operate 24/7. Now, in an attempt to cater to its vast AI needs, Google is going nuclear. The tech giant has signed a deal with California-based nuclear firm Kairos Power to build new nuclear reactors to supply its US data centres with energy. Although the location of these reactors is yet to be revealed, Google said the first will be operational in 2030, with more to follow by 2035.

Google has signed a "world first" deal to buy energy from a fleet of mini nuclear reactors to generate the power needed for the rise in use of artificial intelligence. The US tech corporation has ordered six or seven small nuclear reactors (SMRs) from California's Kairos Power, with the first due to be completed by 2030 and the remainder by 2035. Google hopes the deal will provide a low carbon solution to power datacentres, which require huge volumes of electricity. The US company, owned by Alphabet, said nuclear provided "a clean, round-the-clock power source that can help us reliably meet electricity demands". The explosive growth of generative AI, as well as cloud storage, has increased tech companies' electricity demands.

"The grid needs new electricity sources to support AI technologies," said Michael Terrell, senior director for energy and climate at Google. "This agreement helps accelerate a new technology to meet energy needs cleanly and reliably, and unlock the full potential of AI for everyone." The deal with Google "is important to accelerate the commercialisation of advanced nuclear energy by demonstrating the technical and market viability of a solution critical to decarbonising power grids," said Kairos executive Jeff Olson. The plans still have to be approved by the US Nuclear Regulatory Commission as well as local agencies before they are allowed to proceed. Last year, US regulators gave California-based Kairos Power the first permit in 50 years to build a new type of nuclear reactor.