This paper provides a comprehensive review of mechanical equipment fault diagnosis methods, focusing on the advancements brought by Transformer-based models. It details the structure, working principles, and benefits of Transformers, particularly their self-attention mechanism and parallel computation capabilities, which have propelled their widespread application in natural language processing and computer vision. The discussion highlights key Transformer model variants, such as Vision Transformers (ViT) and their extensions, which leverage self-attention to improve accuracy and efficiency in visual tasks. Furthermore, the paper examines the application of Transformer-based approaches in intelligent fault diagnosis for mechanical systems, showcasing their superior ability to extract and recognize patterns from complex sensor data for precise fault identification. Despite these advancements, challenges remain, including the reliance on extensive labeled datasets, significant computational demands, and difficulties in deploying models on resource-limited devices. To address these limitations, the paper proposes future research directions, such as developing lightweight Transformer architectures, integrating multimodal data sources, and enhancing adaptability to diverse operational conditions. These efforts aim to further expand the application of Transformer-based methods in mechanical fault diagnosis, making them more robust, efficient, and suitable for real-world industrial environments.

Interpretability methods seek to understand language model representations, yet the outputs of most such methods -- circuits, vectors, scalars -- are not immediately human-interpretable. In response, we introduce LatentQA, the task of answering open-ended questions about model activations in natural language. Towards solving LatentQA, we propose Latent Interpretation Tuning (LIT), which finetunes a decoder LLM on a dataset of activations and associated question-answer pairs, similar to how visual instruction tuning trains on question-answer pairs associated with images. We use the decoder for diverse reading applications, such as extracting relational knowledge from representations or uncovering system prompts governing model behavior. Our decoder also specifies a differentiable loss that we use to control models, such as debiasing models on stereotyped sentences and controlling the sentiment of generations. Finally, we extend LatentQA to reveal harmful model capabilities, such as generating recipes for bioweapons and code for hacking.

We study the problem of learning a $k$-body Hamiltonian with $M$ unknown Pauli terms that are not necessarily geometrically local. We propose a protocol that learns the Hamiltonian to precision $\epsilon$ with total evolution time ${\mathcal{O}}(M^{1/2+1/p}/\epsilon)$ up to logarithmic factors, where the error is quantified by the $\ell^p$-distance between Pauli coefficients. Our learning protocol uses only single-qubit control operations and a GHZ state initial state, is non-adaptive, is robust against SPAM errors, and performs well even if $M$ and $k$ are not precisely known in advance or if the Hamiltonian is not exactly $M$-sparse. Methods from the classical theory of compressed sensing are used for efficiently identifying the $M$ terms in the Hamiltonian from among all possible $k$-body Pauli operators. We also provide a lower bound on the total evolution time needed in this learning task, and we discuss the operational interpretations of the $\ell^1$ and $\ell^2$ error metrics. In contrast to most previous works, our learning protocol requires neither geometric locality nor any other relaxed locality conditions.

As the global demand for clean energy intensifies to achieve sustainability and net-zero carbon emission goals, nuclear energy stands out as a reliable solution. However, fully harnessing its potential requires overcoming key challenges, such as the high capital costs associated with nuclear power plants (NPPs). One promising strategy to mitigate these costs involves repurposing sites with existing infrastructure, including coal power plant (CPP) locations, which offer pre-built facilities and utilities. Additionally, brownfield sites - previously developed or underutilized lands often impacted by industrial activity - present another compelling alternative. These sites typically feature valuable infrastructure that can significantly reduce the costs of NPP development. This study introduces a novel multi-objective optimization methodology, leveraging combinatorial search to evaluate over 30,000 potential NPP sites in the United States. Our approach addresses gaps in the current practice of assigning pre-determined weights to each site attribute that could lead to bias in the ranking. Each site is assigned a performance-based score, derived from a detailed combinatorial analysis of its site attributes. The methodology generates a comprehensive database comprising site locations (inputs), attributes (outputs), site score (outputs), and the contribution of each attribute to the site score (outputs). We then use this database to train a machine learning neural network model, enabling rapid predictions of nuclear siting suitability across any location in the contiguous United States.

Energy efficiency of Convolutional Neural Networks (CNNs) has become an important area of research, with various strategies being developed to minimize the power consumption of these models. Previous efforts, including techniques like model pruning, quantization, and hardware optimization, have made significant strides in this direction. However, there remains a need for more effective on device AI solutions that balance energy efficiency with model performance. In this paper, we propose a novel approach to reduce the energy requirements of inference of CNNs. Our methodology employs two small Complementary CNNs that collaborate with each other by covering each other's "weaknesses" in predictions. If the confidence for a prediction of the first CNN is considered low, the second CNN is invoked with the aim of producing a higher confidence prediction. This dual-CNN setup significantly reduces energy consumption compared to using a single large deep CNN. Additionally, we propose a memory component that retains previous classifications for identical inputs, bypassing the need to re-invoke the CNNs for the same input, further saving energy. Our experiments on a Jetson Nano computer demonstrate an energy reduction of up to 85.8% achieved on modified datasets where each sample was duplicated once. These findings indicate that leveraging a complementary CNN pair along with a memory component effectively reduces inference energy while maintaining high accuracy.

Concerns about the environmental footprint of machine learning are increasing. While studies of energy use and emissions of ML models are a growing subfield, most ML researchers and developers still do not incorporate energy measurement as part of their work practices. While measuring energy is a crucial step towards reducing carbon footprint, it is also not straightforward. This paper introduces the main considerations necessary for making sound use of energy measurement tools and interpreting energy estimates, including the use of at-the-wall versus on-device measurements, sampling strategies and best practices, common sources of error, and proxy measures. It also contains practical tips and real-world scenarios that illustrate how these considerations come into play. It concludes with a call to action for improving the state of the art of measurement methods and standards for facilitating robust comparisons between diverse hardware and software environments.

As data centres consume even more energy to serve the intensive computing needs of artificial intelligence, they could contribute to an estimated 600,000 asthma cases and 1300 premature deaths per year by 2030 – accounting for more than one third of asthma deaths annually in the US. "Public health impacts are direct and tangible impacts on people, and these impacts are substantial and not limited to a small radius of where data centres operate," says Shaolei Ren at the University of California, Riverside. "They affect people across the country." Ren and his colleagues, including Adam Wierman at the California Institute of Technology, developed those estimates based on data centres' projected electricity demand, which produces additional emissions and contributes to air pollution. For instance, the electricity usage required for training large AI models could produce air pollutants equivalent to driving a passenger car for more than 10,000 roundtrips between Los Angeles and New York City, according to the researchers.

A Ukrainian video game is shattering records since its release on November 20, despite many delays and its being a target of a Russian disinformation operation. S.T.A.L.K.E.R. 2: Heart of Chornobyl is the latest edition of a game series that started in 2007 with S.T.A.L.K.E.R.: Shadow of Chornobyl, developed by GSC Game World, a Ukrainian video games studio. The game surpassed one million downloads and 117,000 concurrent players within 48 hours of its release, making it the most successful Ukrainian-developed title to date. Yet, that landmark achievement in the country's gaming industry was bittersweet. Former GSC Game World developer Volodymyr Yezhov, who worked on S.T.A.L.K.E.R. 2 and was known by the nickname "Fresh", was killed in combat near Bakhmut in December 2022, while serving in the Ukrainian military.

Tesla lobbied the UK government to strengthen rules on carbon emissions from cars and lorries, according to documents that also show the electric carmaker continued to push for increased taxes on fossil fuel cars. The US carmaker, which is run by Elon Musk, pushed for the British government to strengthen its zero-emission vehicle (ZEV) mandate for cars and introduce equivalent rules for heavy goods vehicles (HGVs), in a letter to Lilian Greenwood, the Labour roads minister. Musk has launched a public feud with Labour, but his company has been more complimentary. A Tesla vice-president wrote in July that "we applaud the Labour party's strong position to decarbonisation of the energy system by 2030, growth and net zero". The letter was obtained under freedom of information laws by the Fast Charge newsletter and shared with the Guardian.

First principles models, derived from fundamental physical laws, have been instrumental in the development of scientific theories and technological systems. For example, the Navier-Stokes equation offers a comprehensive description of fluid flow, enabling predictions of complex behaviors in everything from blood flow [1] to weather patterns [2]. Traditionally, this pursuit has relied on the extensive expertise of domain specialists, requiring trial and error to identify features and model structures that fit the observations. In recent years, the landscape of scientific inquiry has been transformed by the availability of machine learning frameworks, such as neural networks, support vector machines, and Gaussian processes, which offer a powerful alternative for deriving predictive models [3]. These data-driven regression methods are often complex, do not typically generalize outside of the training set, and provide limited insights into the underlying physics. For instance, while these models may be trained to accurately predict the Reynolds number, they cannot capture the competitive nature between inertial and viscous forces in fluid flow. The only data-driven modeling framework that can provide insights comparable to first principles models, to the best of our knowledge, is symbolic regression (SR) [4, 5, 6].