Search

The rapid development of large language models (LLMs) has significantly transformed the field of artificial intelligence, demonstrating remarkable capabilities in natural language processing and moving towards multi-modal functionality. These models are increasingly integrated into diverse applications, impacting both research and industry. However, their development and deployment present substantial challenges, including the need for extensive computational resources, high energy consumption, and complex software optimizations. Unlike traditional deep learning systems, LLMs require unique optimization strategies for training and inference, focusing on system-level efficiency. This paper surveys hardware and software co-design approaches specifically tailored to address the unique characteristics and constraints of large language models. This survey analyzes the challenges and impacts of LLMs on hardware and algorithm research, exploring algorithm optimization, hardware design, and system-level innovations. It aims to provide a comprehensive understanding of the trade-offs and considerations in LLM-centric computing systems, guiding future advancements in AI. Finally, we summarize the existing efforts in this space and outline future directions toward realizing production-grade co-design methodologies for the next generation of large language models and AI systems., Comment: Accepted by IEEE Circuits and Systems Magazine

Quantum discord can demonstrate quantum nonlocality in the context of a semi-device-independent Bell or steering scenario, i.e., by assuming only the Hilbert-space dimension. This work addresses which aspect of bipartite coherence is essential to such semi-device-independent quantum information tasks going beyond standard Bell nonlocality or quantum steering. It has been shown that the global coherence of a single system can be transformed into bipartite entanglement. However, global coherence can also be present in quantum discord. At the same time, discord can display bipartite coherence locally, i.e., only in a subsystem or both subsystems. Thus, global coherence of bipartite separable states is defined here as a form of bipartite coherence that is not reducible to local coherence in any of the subsystems or both subsystems. To operationally identify such a notion of global coherence in discord, we demonstrate that global coherence is necessary to demonstrate semi-device-independent nonlocality of quantum discord in Bell or steering scenarios. From this result, it follows that any local operations of the form $\Phi_A \otimes \Phi_B$ that may create coherence locally are free operations in the resource theory of semi-device-independent nonlocality of discord. Applying this resource theory, it is identified that the quantum nonlocality beyond quantum steering acts as a resource for the quantum communication task of remote state preparation using two-qubit separable states., Comment: 12 pages, 4 figures

Monolithic liquid scintillator detector technology is the workhorse for detecting neutrinos and exploring new physics. The KamLAND-Zen experiment exemplifies this detector technology and has yielded top results in the quest for neutrinoless double-beta ($0\nu\beta\beta$) decay. To understand the physical events that occur in the detector, experimenters must reconstruct each event's position and energy from the raw data produced. Traditionally, this information has been obtained through a time-consuming offline process, meaning that event position and energy would only be available days after data collection. This work introduces a new pipeline to acquire this information quickly by implementing a machine learning model, PointNet, onto a Field Programmable Gate Array (FPGA). This work outlines a successful demonstration of the entire pipeline, showing that event position and energy information can be reliably and quickly obtained as physics events occur in the detector. This marks one of the first instances of applying hardware-AI co-design in the context of $0\nu\beta\beta$ decay experiments.

The total variation (TV) method is an image denoising technique that aims to reduce noise by minimizing the total variation of the image, which measures the variation in pixel intensities. The TV method has been widely applied in image processing and computer vision for its ability to preserve edges and enhance image quality. In this paper, we propose an improved TV model for image denoising and the associated numerical algorithm to carry out the procedure, which is particularly effective in removing several types of noises and their combinations. Our improved model admits a unique solution and the associated numerical algorithm guarantees the convergence. Numerical experiments are demonstrated to show improved effectiveness and denoising quality compared to other TV models. Such encouraging results further enhance the utility of the TV method in image processing.

Arthroscopy is a minimally invasive surgical procedure used to diagnose and treat joint problems. The clinical workflow of arthroscopy typically involves inserting an arthroscope into the joint through a small incision, during which surgeons navigate and operate largely by relying on their visual assessment through the arthroscope. However, the arthroscope's restricted field of view and lack of depth perception pose challenges in navigating complex articular structures and achieving surgical precision during procedures. Aiming at enhancing intraoperative awareness, we present a robust pipeline that incorporates simultaneous localization and mapping, depth estimation, and 3D Gaussian splatting to realistically reconstruct intra-articular structures solely based on monocular arthroscope video. Extending 3D reconstruction to Augmented Reality (AR) applications, our solution offers AR assistance for articular notch measurement and annotation anchoring in a human-in-the-loop manner. Compared to traditional Structure-from-Motion and Neural Radiance Field-based methods, our pipeline achieves dense 3D reconstruction and competitive rendering fidelity with explicit 3D representation in 7 minutes on average. When evaluated on four phantom datasets, our method achieves RMSE = 2.21mm reconstruction error, PSNR = 32.86 and SSIM = 0.89 on average. Because our pipeline enables AR reconstruction and guidance directly from monocular arthroscopy without any additional data and/or hardware, our solution may hold the potential for enhancing intraoperative awareness and facilitating surgical precision in arthroscopy. Our AR measurement tool achieves accuracy within 1.59 +/- 1.81mm and the AR annotation tool achieves a mIoU of 0.721., Comment: 8 pages, with 2 additional pages as the supplementary. Accepted by AE-CAI 2024

In percutaneous pelvic trauma surgery, accurate placement of Kirschner wires (K-wires) is crucial to ensure effective fracture fixation and avoid complications due to breaching the cortical bone along an unsuitable trajectory. Surgical navigation via mixed reality (MR) can help achieve precise wire placement in a low-profile form factor. Current approaches in this domain are as yet unsuitable for real-world deployment because they fall short of guaranteeing accurate visual feedback due to uncontrolled bending of the wire. To ensure accurate feedback, we introduce StraightTrack, an MR navigation system designed for percutaneous wire placement in complex anatomy. StraightTrack features a marker body equipped with a rigid access cannula that mitigates wire bending due to interactions with soft tissue and a covered bony surface. Integrated with an Optical See-Through Head-Mounted Display (OST HMD) capable of tracking the cannula body, StraightTrack offers real-time 3D visualization and guidance without external trackers, which are prone to losing line-of-sight. In phantom experiments with two experienced orthopedic surgeons, StraightTrack improves wire placement accuracy, achieving the ideal trajectory within $5.26 \pm 2.29$ mm and $2.88 \pm 1.49$ degree, compared to over 12.08 mm and 4.07 degree for comparable methods. As MR navigation systems continue to mature, StraightTrack realizes their potential for internal fracture fixation and other percutaneous orthopedic procedures.

In this work, we propose a novel consistency-preserving loss function for recovering the phase information in the context of phase reconstruction (PR) and speech enhancement (SE). Different from conventional techniques that directly estimate the phase using a deep model, our idea is to exploit ad-hoc constraints to directly generate a consistent pair of magnitude and phase. Specifically, the proposed loss forces a set of complex numbers to be a consistent short-time Fourier transform (STFT) representation, i.e., to be the spectrogram of a real signal. Our approach thus avoids the difficulty of estimating the original phase, which is highly unstructured and sensitive to time shift. The influence of our proposed loss is first assessed on a PR task, experimentally demonstrating that our approach is viable. Next, we show its effectiveness on an SE task, using both the VB-DMD and WSJ0-CHiME3 data sets. On VB-DMD, our approach is competitive with conventional solutions. On the challenging WSJ0-CHiME3 set, the proposed framework compares favourably over those techniques that explicitly estimate the phase., Comment: 5 pages, Submitted to ICASSP 2025

This paper proposes a generative pretraining foundation model for high-quality speech restoration tasks. By directly operating on complex-valued short-time Fourier transform coefficients, our model does not rely on any vocoders for time-domain signal reconstruction. As a result, our model simplifies the synthesis process and removes the quality upper-bound introduced by any mel-spectrogram vocoder compared to prior work SpeechFlow. The proposed method is evaluated on multiple speech restoration tasks, including speech denoising, bandwidth extension, codec artifact removal, and target speaker extraction. In all scenarios, finetuning our pretrained model results in superior performance over strong baselines. Notably, in the target speaker extraction task, our model outperforms existing systems, including those leveraging SSL-pretrained encoders like WavLM. The code and the pretrained checkpoints are publicly available in the NVIDIA NeMo framework., Comment: 5 pages, Submitted to ICASSP 2025. The implementation and configuration could be found in https://github.com/NVIDIA/NeMo/blob/main/examples/audio/conf/flow_matching_generative_ssl_pretraining.yaml The audio demo page could be found in https://kuray107.github.io/ssl_gen25-examples/index.html

We present OpenDosimeter (https://opendosimeter.org/), an open hardware solution for real-time personal X-ray dose monitoring based on a scintillation counter. Using an X-ray sensor assembly (LYSO + SiPM) on a custom board powered by a Raspberry Pi Pico, OpenDosimeter provides real-time feedback (1 Hz), data logging (10 hours), and battery-powered operation. One of the core innovations is that we calibrate the device using $^{241}$Am found in ionization smoke detectors. Specifically, we use the $\gamma$-emissions to spectrally calibrate the dosimeter, then calculate the effective dose from X-ray exposure by compensating for the scintillator absorption efficiency and applying energy-to-dose coefficients derived from tabulated data in the ICRP 116 publication. We demonstrate that this transparent approach enables real-time dose rate readings with a linear response between 0.1-1000 $\mu$Sv/h at $\pm$25% accuracy, tested for energies up to 120 keV. The maximum dose rate readings are limited by pile-up effects when approaching count rate saturation ($\sim$77 kcps at $\sim$13 $\mu$s average pulse processing time). The total component cost for making an OpenDosimeter is <\$100, which, combined with its open design (both hardware and software), enables cost-effective local reproducibility on a global scale. This paper complements the open-source documentation by explaining the underlying technology, the algorithm for dose calculation, and areas for future improvement., Comment: 7 pages, 3 figures

Mathematical reasoning remains an ongoing challenge for AI models, especially for geometry problems that require both linguistic and visual signals. As the vision encoders of most MLLMs are trained on natural scenes, they often struggle to understand geometric diagrams, performing no better in geometry problem solving than LLMs that only process text. This limitation is amplified by the lack of effective methods for representing geometric relationships. To address these issues, we introduce the Diagram Formalization Enhanced Geometry Problem Solver (DFE-GPS), a new framework that integrates visual features, geometric formal language, and natural language representations. We propose a novel synthetic data approach and create a large-scale geometric dataset, SynthGeo228K, annotated with both formal and natural language captions, designed to enhance the vision encoder for a better understanding of geometric structures. Our framework improves MLLMs' ability to process geometric diagrams and extends their application to open-ended tasks on the formalgeo7k dataset.

Open quantum systems are susceptible to losses in information, energy, and particles due to their surrounding environment. One novel strategy to mitigate these losses is to transform them into advantages for quantum technologies through tailored non-Hermitian quantum systems. In this work, we theoretically propose a fast generation of multipartite entanglement in non-Hermitian qubits. Our findings reveal that weakly coupled non-Hermitian qubits can accelerate multiparty entanglement generation by thousands of times compared to Hermitian qubits, in particular when approaching the $2^n$-th order exceptional points of $n$ qubits in the ${\cal P}{\cal T}-$ symmetric regime. Furthermore, we show that Hermitian qubits can generate GHZ states with a high fidelity more than $0.9995$ in a timescale comparable to that of non-Hermitian qubits, but at the expense of intense driving and large coupling constant. Our approach is scalable to a large number of qubits, presenting a promising pathway for advancing quantum technologies through the non-Hermiticity and higher-order exceptional points in many-body quantum systems., Comment: 5 figures

Machine learning models are often criticized for their black-box nature, raising concerns about their applicability in critical decision-making scenarios. Consequently, there is a growing demand for interpretable models in such contexts. In this study, we introduce Model-based Deep Rule Forests (mobDRF), an interpretable representation learning algorithm designed to extract transparent models from data. By leveraging IF-THEN rules with multi-level logic expressions, mobDRF enhances the interpretability of existing models without compromising accuracy. We apply mobDRF to identify key risk factors for cognitive decline in an elderly population, demonstrating its effectiveness in subgroup analysis and local model optimization. Our method offers a promising solution for developing trustworthy and interpretable machine learning models, particularly valuable in fields like healthcare, where understanding differential effects across patient subgroups can lead to more personalized and effective treatments.

The Special Euclidean group on the plane SE(2) has the left-invariant sub- Riemannian structure. Every sub-Riemannian manifold possesses a Hamiltonian function governing the sub-Riemannian geodesic flow. Two natural questions are: What are the necessary conditions for periodic sub-Riemannian geodesics? What geodesics are the metric lines in SE(2)? We answer both questions, and our method for the second is an alternative proof using the Hamilton-Jacobi theory.

The uncertainty principle is one of quantum theory's most foundational features. It underpins a quantum phenomenon called measurement incompatibility -- two physical observables of a single quantum system may not always be measured simultaneously. Apart from being fundamentally important, measurement incompatibility is also a powerful resource in the broad quantum science and technologies, with wide applications to cryptography, communication, random number generation, and device-independent tasks. Since every physical system is unavoidably subject to noise, an important, yet still open, question is how to characterise the ability of noisy quantum dynamics to preserve measurement incompatibility. This work fills this gap by providing the first resource theory of this ability, termed incompatibility preservability. We quantify incompatibility preservability by a robustness measure. Then, we introduce an operational task, entanglement-assisted filter game, to completely characterise both the robustness measure and the conversion of incompatibility preservability. Our results provide a general framework to describe how noisy dynamics affect the uncertainty principle's signature., Comment: 4+5 pages; 4 figures

Geometric frustration is known to completely damage kinetic processes of some of the orbitals (and their associated quantum coherence) as to produce flat bands in the non-interacting systems. The impact of introducing additional interaction to the system in such frustrated systems is, however, a highly controversial issue. On the one hand, numerical studies on geometrically frustrated systems of hard-core boson (equivalent to a spin-1/2 systems) typically lead to glass or solid phases containing only local many-body coherence, indicating the persistence of the damage in quantum coherence. On the other, there continues to be noticeable claims of development of superfluidity that implies kinetic flow of particles. To resolve this apparent contradiction of great significance, we present a rigorous proof showing that density-density interaction is incapable of defeating the geometric frustration to allow propagation of those immobile particles, let alone sustaining a superfluidity. Instead, the frustrated systems develop many $\textit{compact}$ many-body localized states as "many-body scars" that do not thermalize, making them good candidates for storing $\textit{robust}$ and $\textit{manipulable}$ quantum information., Comment: 6 pages, 4 figures

The reconstruction of electrical current densities from magnetic field measurements is an important technique with applications in materials science, circuit design, quality control, plasma physics, and biology. Analytic reconstruction methods exist for planar currents, but break down in the presence of high spatial frequency noise or large standoff distance, restricting the types of systems that can be studied. Here, we demonstrate the use of a deep convolutional neural network for current density reconstruction from two-dimensional (2D) images of vector magnetic fields acquired by a quantum diamond microscope (QDM) utilizing a surface layer of Nitrogen Vacancy (NV) centers in diamond. Trained network performance significantly exceeds analytic reconstruction for data with high noise or large standoff distances. This machine learning technique can perform quality inversions on lower SNR data, reducing the data collection time by a factor of about 400 and permitting reconstructions of weaker and three-dimensional current sources., Comment: 17 pages, 10 figures. Includes Supplemental Information

Sparse RGBD scene completion is a challenging task especially when considering consistent textures and geometries throughout the entire scene. Different from existing solutions that rely on human-designed text prompts or predefined camera trajectories, we propose GenRC, an automated training-free pipeline to complete a room-scale 3D mesh with high-fidelity textures. To achieve this, we first project the sparse RGBD images to a highly incomplete 3D mesh. Instead of iteratively generating novel views to fill in the void, we utilized our proposed E-Diffusion to generate a view-consistent panoramic RGBD image which ensures global geometry and appearance consistency. Furthermore, we maintain the input-output scene stylistic consistency through textual inversion to replace human-designed text prompts. To bridge the domain gap among datasets, E-Diffusion leverages models trained on large-scale datasets to generate diverse appearances. GenRC outperforms state-of-the-art methods under most appearance and geometric metrics on ScanNet and ARKitScenes datasets, even though GenRC is not trained on these datasets nor using predefined camera trajectories. Project page: https://minfenli.github.io/GenRC, Comment: ECCV 2024

A method for assessing the quality of electronic material properties of thin-film metal oxide semiconductor field-effect transistors (MOSFETs) is presented. By investigating samples with MOCVD-grown MoS${_2}$ channels exposed to atmospheric conditions, the existence of electron traps in MoS${_2}$ and at the interface between the gate insulator and the thin-film MoS${_2}$ are revealed. Differential conductance and capacitance data of the transistor channels are plotted as 3D surfaces on a base plane spanned by the measurement frequency versus the gate voltage. The existence of defects is confirmed by comparison with ideal results from a theoretical model., Comment: 24 pages

Strongly correlated materials are known to display qualitatively distinct emergent behaviors at low energy. Conveniently, the superposition principle of quantum mechanics ensures that, upon absorbing quantum fluctuation, these rich low-energy behaviors can always be effectively described by dressed particles with fully quantized charge, spin, and orbitals structure. Such a powerful and simple description is, however, difficult to access through density functional theory (DFT) calculations, since in terms of bare particles the quantum fluctuation would heavily smear the quantized quantities. To address this difficulty, we propose an `interaction annealing' approach to decipher the dominant valence and orbital structure by suppressing the charge fluctuation through enhancing ionic charging energy. Applying this approach to ferroelectric semi-metal WTe${_2}$ as a demonstration, we identify a dominant ferro-orbital ordered structure with W ion in a $d^2$ spin-0 configuration. The proposed approach is straightforward to implement in standard DFT calculations to grant additional access to essential low-energy physics., Comment: 6 pages, 3 figures

Multiple types of inference are available for probabilistic graphical models, e.g., marginal, maximum-a-posteriori, and even marginal maximum-a-posteriori. Which one do researchers mean when they talk about "planning as inference"? There is no consistency in the literature, different types are used, and their ability to do planning is further entangled with specific approximations or additional constraints. In this work we use the variational framework to show that all commonly used types of inference correspond to different weightings of the entropy terms in the variational problem, and that planning corresponds _exactly_ to a _different_ set of weights. This means that all the tricks of variational inference are readily applicable to planning. We develop an analogue of loopy belief propagation that allows us to perform approximate planning in factored state Markov decisions processes without incurring intractability due to the exponentially large state space. The variational perspective shows that the previous types of inference for planning are only adequate in environments with low stochasticity, and allows us to characterize each type by its own merits, disentangling the type of inference from the additional approximations that its practical use requires. We validate these results empirically on synthetic MDPs and tasks posed in the International Planning Competition., Comment: Corrected fontsize in Fig. 1. No other changes

The recent years have witnessed great advances in video generation. However, the development of automatic video metrics is lagging significantly behind. None of the existing metric is able to provide reliable scores over generated videos. The main barrier is the lack of large-scale human-annotated dataset. In this paper, we release VideoFeedback, the first large-scale dataset containing human-provided multi-aspect score over 37.6K synthesized videos from 11 existing video generative models. We train VideoScore (initialized from Mantis) based on VideoFeedback to enable automatic video quality assessment. Experiments show that the Spearman correlation between VideoScore and humans can reach 77.1 on VideoFeedback-test, beating the prior best metrics by about 50 points. Further result on other held-out EvalCrafter, GenAI-Bench, and VBench show that VideoScore has consistently much higher correlation with human judges than other metrics. Due to these results, we believe VideoScore can serve as a great proxy for human raters to (1) rate different video models to track progress (2) simulate fine-grained human feedback in Reinforcement Learning with Human Feedback (RLHF) to improve current video generation models.

Goncharov proved an explicit formula for the coproduct in the Hopf algebra of motivic iterated integrals. Yamamoto introduced Yamamoto's integral which generalizes iterated integrals and gave a new integral expression for multiple zeta star values using Yamamoto's integral. In this paper, we consider the motivic version of Yamamoto's integral and generalize Goncharov's coproduct formula to those motivic integrals. As an example, we will compute the coproduct of a certain type of Schur multiple zeta values., Comment: 19 pages

Generative AI has made remarkable strides to revolutionize fields such as image and video generation. These advancements are driven by innovative algorithms, architecture, and data. However, the rapid proliferation of generative models has highlighted a critical gap: the absence of trustworthy evaluation metrics. Current automatic assessments such as FID, CLIP, FVD, etc often fail to capture the nuanced quality and user satisfaction associated with generative outputs. This paper proposes an open platform GenAI-Arena to evaluate different image and video generative models, where users can actively participate in evaluating these models. By leveraging collective user feedback and votes, GenAI-Arena aims to provide a more democratic and accurate measure of model performance. It covers three arenas for text-to-image generation, text-to-video generation, and image editing respectively. Currently, we cover a total of 27 open-source generative models. GenAI-Arena has been operating for four months, amassing over 6000 votes from the community. We describe our platform, analyze the data, and explain the statistical methods for ranking the models. To further promote the research in building model-based evaluation metrics, we release a cleaned version of our preference data for the three tasks, namely GenAI-Bench. We prompt the existing multi-modal models like Gemini, GPT-4o to mimic human voting. We compute the correlation between model voting with human voting to understand their judging abilities. Our results show existing multimodal models are still lagging in assessing the generated visual content, even the best model GPT-4o only achieves a Pearson correlation of 0.22 in the quality subscore, and behaves like random guessing in others., Comment: 9 pages,7 figures

We propose a novel language-universal approach to end-to-end automatic spoken keyword recognition (SKR) leveraging upon (i) a self-supervised pre-trained model, and (ii) a set of universal speech attributes (manner and place of articulation). Specifically, Wav2Vec2.0 is used to generate robust speech representations, followed by a linear output layer to produce attribute sequences. A non-trainable pronunciation model then maps sequences of attributes into spoken keywords in a multilingual setting. Experiments on the Multilingual Spoken Words Corpus show comparable performances to character- and phoneme-based SKR in seen languages. The inclusion of domain adversarial training (DAT) improves the proposed framework, outperforming both character- and phoneme-based SKR approaches with 13.73% and 17.22% relative word error rate (WER) reduction in seen languages, and achieves 32.14% and 19.92% WER reduction for unseen languages in zero-shot settings.

In the age of large-scale language models, benchmarks like the Massive Multitask Language Understanding (MMLU) have been pivotal in pushing the boundaries of what AI can achieve in language comprehension and reasoning across diverse domains. However, as models continue to improve, their performance on these benchmarks has begun to plateau, making it increasingly difficult to discern differences in model capabilities. This paper introduces MMLU-Pro, an enhanced dataset designed to extend the mostly knowledge-driven MMLU benchmark by integrating more challenging, reasoning-focused questions and expanding the choice set from four to ten options. Additionally, MMLU-Pro eliminates the trivial and noisy questions in MMLU. Our experimental results show that MMLU-Pro not only raises the challenge, causing a significant drop in accuracy by 16% to 33% compared to MMLU but also demonstrates greater stability under varying prompts. With 24 different prompt styles tested, the sensitivity of model scores to prompt variations decreased from 4-5% in MMLU to just 2% in MMLU-Pro. Additionally, we found that models utilizing Chain of Thought (CoT) reasoning achieved better performance on MMLU-Pro compared to direct answering, which is in stark contrast to the findings on the original MMLU, indicating that MMLU-Pro includes more complex reasoning questions. Our assessments confirm that MMLU-Pro is a more discriminative benchmark to better track progress in the field., Comment: This version has been accepted and published at NeurIPS 2024 Track Datasets and Benchmarks (Spotlight)

This paper provides an overview of the Fake-EmoReact 2021 Challenge, held at the 9th SocialNLP Workshop, in conjunction with NAACL 2021. The challenge requires predicting the authenticity of tweets using reply context and augmented GIF categories from EmotionGIF dataset. We offer the Fake-EmoReact dataset with more than 453k as the experimental materials, where every tweet is labeled with authenticity. Twenty-four teams registered to participate in this challenge, and 5 submitted their results successfully in the evaluation phase. The best team achieves 93.9 on Fake-EmoReact 2021 dataset using F1 score. In addition, we show the definition of share task, data collection, and the teams' performance that joined this challenge and their approaches.

We study the essential self-adjointness of semi-bounded Schr\"{o}dinger operators on birth-death chains. First, we offer a general characterization which originates from studying a second order linear recurrence with variational coefficients which comes from the Schr\"{o}dinger operator. As this characterization is algebraically complicated, we present an additional theorem discussing the failure of essential self-adjointness. Finally, we study two specific cases of solutions to equations involving the Schr\"{o}dinger operator over birth-death chains and derive explicit formulas in these cases., Comment: 13 pages

Energy storage devices, such as batteries, thermal energy storages, and hydrogen systems, can help mitigate climate change by ensuring a more stable and sustainable power supply. To maximize the effectiveness of such energy storage, determining the appropriate charging and discharging amounts for each time period is crucial. Reinforcement learning is preferred over traditional optimization for the control of energy storage due to its ability to adapt to dynamic and complex environments. However, the continuous nature of charging and discharging levels in energy storage poses limitations for discrete reinforcement learning, and time-varying feasible charge-discharge range based on state of charge (SoC) variability also limits the conventional continuous reinforcement learning. In this paper, we propose a continuous reinforcement learning approach that takes into account the time-varying feasible charge-discharge range. An additional objective function was introduced for learning the feasible action range for each time period, supplementing the objectives of training the actor for policy learning and the critic for value learning. This actively promotes the utilization of energy storage by preventing them from getting stuck in suboptimal states, such as continuous full charging or discharging. This is achieved through the enforcement of the charging and discharging levels into the feasible action range. The experimental results demonstrated that the proposed method further maximized the effectiveness of energy storage by actively enhancing its utilization., Comment: ICLR 2024 Workshop: Tackling Climate Change with Machine Learning

In economic program evaluation, it is common to obtain panel data in which outcomes are indicators that an individual has reached an absorbing state. For example, they may indicate whether an individual has exited a period of unemployment, passed an exam, left a marriage, or had their parole revoked. The parallel trends assumption that underpins difference-in-differences generally fails in such settings. We suggest identifying conditions that are analogous to those of difference-in-differences but apply to hazard rates rather than mean outcomes. These alternative assumptions motivate estimators that retain the simplicity and transparency of standard diff-in-diff, and we suggest analogous specification tests. Our approach can be adapted to general linear restrictions between the hazard rates of different groups, motivating duration analogues of the triple differences and synthetic control methods. We apply our procedures to examine the impact of a policy that increased the generosity of unemployment benefits, using a cross-cohort comparison., Comment: 46 pages, 7 figures, 2 tables

Large multimodal models (LMMs) have shown great results in single-image vision language tasks. However, their abilities to solve multi-image visual language tasks is yet to be improved. The existing LMMs like OpenFlamingo, Emu2, Idefics gain their multi-image ability through pre-training on hundreds of millions of noisy interleaved image-text data from the web, which is neither efficient nor effective. In this paper, we aim to build strong multi-image LMMs via instruction tuning with academic-level resources. Therefore, we meticulously construct Mantis-Instruct containing 721K multi-image instruction data to train a family of models Mantis. The instruction tuning empowers Mantis with different multi-image skills like co-reference, comparison, reasoning, and temporal understanding. We evaluate Mantis on five multi-image benchmarks and seven single-image benchmarks. Mantis-SigLIP can achieve SoTA results on all the multi-image benchmarks and beat the strongest multi-image baseline, Idefics2-8B by an average of 11 absolute points. Notably, Idefics2-8B was pre-trained on 140M interleaved multi-image data, which is 200x larger than Mantis-Instruct. We observe that Mantis performs equivalently well on the held-in and held-out benchmarks, which shows its generalization ability. Notably, we found that Mantis can even match the performance of GPT-4V on multi-image benchmarks. We further evaluate Mantis on single-image benchmarks and demonstrate that Mantis also maintains a strong single-image performance on par with CogVLM and Emu2. Our results show that multi-image abilities are not necessarily gained through massive pre-training, instead, it can be gained by the low-cost instruction tuning. Our work provides new perspectives on how to improve LMMs' multi-image abilities., Comment: 9 pages, 3 figures, 8 tables