Large language models (LLMs) have achieved huge success for their general knowledge and ability to solve a wide spectrum of tasks in natural language processing (NLP). Due to their impressive abilities, LLMs have shed light on potential inter-discipline applications to foster scientific discoveries of a specific domain by using artificial intelligence (AI for science, AI4S). In the meantime, utilizing NLP techniques in geoscience research and practice is wide and convoluted, contributing from knowledge extraction and document classification to question answering and knowledge discovery. In this work, we take the initial step to leverage LLM for science, through a rather straightforward approach. We try to specialize an LLM into geoscience, by further pre-training the model with a vast amount of texts in geoscience, as well as supervised fine-tuning (SFT) the resulting model with our custom collected instruction tuning dataset. These efforts result in a model GeoGalactica consisting of 30 billion parameters. To our best knowledge, it is the largest language model for the geoscience domain. More specifically, GeoGalactica is from further pre-training of Galactica. We train GeoGalactica over a geoscience-related text corpus containing 65 billion tokens, preserving as the largest geoscience-specific text corpus. Then we fine-tune the model with 1 million pairs of instruction-tuning data consisting of questions that demand professional geoscience knowledge to answer. In this technical report, we will illustrate in detail all aspects of GeoGalactica, including data collection, data cleaning, base model selection, pre-training, SFT, and evaluation. We open-source our data curation tools and the checkpoints of GeoGalactica during the first 3/4 of pre-training.
Structured data in the form of tabular datasets contain features that are distinct and discrete, with varying individual and relative importances to the target. Combinations of one or more features may be more predictive and meaningful than simple individual feature contributions. R's mixed effect linear models library allows users to provide such interactive feature combinations in the model design. However, given many features and possible interactions to select from, model selection becomes an exponentially difficult task. We aim to automate the model selection process for predictions on tabular datasets incorporating feature interactions while keeping computational costs small. The framework includes two distinct approaches for feature selection: a Priority-based Random Grid Search and a Greedy Search method. The Priority-based approach efficiently explores feature combinations using prior probabilities to guide the search. The Greedy method builds the solution iteratively by adding or removing features based on their impact. Experiments on synthetic demonstrate the ability to effectively capture predictive feature combinations.
Neural marked temporal point processes have been a valuable addition to the existing toolbox of statistical parametric models for continuous-time event data. These models are useful for sequences where each event is associated with a single item (a single type of event or a "mark") -- but such models are not suited for the practical situation where each event is associated with a set of items. In this work, we develop a general framework for modeling set-valued data in continuous-time, compatible with any intensity-based recurrent neural point process model. In addition, we develop inference methods that can use such models to answer probabilistic queries such as "the probability of item $A$ being observed before item $B$," conditioned on sequence history. Computing exact answers for such queries is generally intractable for neural models due to both the continuous-time nature of the problem setting and the combinatorially-large space of potential outcomes for each event. To address this, we develop a class of importance sampling methods for querying with set-based sequences and demonstrate orders-of-magnitude improvements in efficiency over direct sampling via systematic experiments with four real-world datasets. We also illustrate how to use this framework to perform model selection using likelihoods that do not involve one-step-ahead prediction.
Human annotations are vital to supervised learning, yet annotators often disagree on the correct label, especially as annotation tasks increase in complexity. A strategy to improve label quality is to ask multiple annotators to label the same item and aggregate their labels. Many aggregation models have been proposed for categorical or numerical annotation tasks, but far less work has considered more complex annotation tasks involving open-ended, multivariate, or structured responses. While a variety of bespoke models have been proposed for specific tasks, our work is the first to introduce aggregation methods that generalize across many diverse complex tasks, including sequence labeling, translation, syntactic parsing, ranking, bounding boxes, and keypoints. This generality is achieved by devising a task-agnostic method to model distances between labels rather than the labels themselves. This article extends our prior work with investigation of three new research questions. First, how do complex annotation properties impact aggregation accuracy? Second, how should a task owner navigate the many modeling choices to maximize aggregation accuracy? Finally, what diagnoses can verify that aggregation models are specified correctly for the given data? To understand how various factors impact accuracy and to inform model selection, we conduct simulation studies and experiments on real, complex datasets. Regarding testing, we introduce unit tests for aggregation models and present a suite of such tests to ensure that a given model is not mis-specified and exhibits expected behavior. Beyond investigating these research questions above, we discuss the foundational concept of annotation complexity, present a new aggregation model as a bridge between traditional models and our own, and contribute a new semi-supervised learning method for complex label aggregation that outperforms prior work.
Automated machine learning (AutoML) streamlines the creation of ML models. While most methods select the "best" model based on predictive quality, it's crucial to acknowledge other aspects, such as interpretability and resource consumption. This holds particular importance in the context of deep neural networks (DNNs), as these models are often perceived as computationally intensive black boxes. In the challenging domain of time series forecasting, DNNs achieve stunning results, but specialized approaches for automatically selecting models are scarce. In this paper, we propose AutoXPCR - a novel method for automated and explainable multi-objective model selection. Our approach leverages meta-learning to estimate any model's performance along PCR criteria, which encompass (P)redictive error, (C)omplexity, and (R)esource demand. Explainability is addressed on multiple levels, as our interactive framework can prioritize less complex models and provide by-product explanations of recommendations. We demonstrate practical feasibility by deploying AutoXPCR on over 1000 configurations across 114 data sets from various domains. Our method clearly outperforms other model selection approaches - on average, it only requires 20% of computation costs for recommending models with 90% of the best-possible quality.
The recent explosion of interest in multimodal applications has resulted in a wide selection of datasets and methods for representing and integrating information from different modalities. Despite these empirical advances, there remain fundamental research questions: How can we quantify the interactions that are necessary to solve a multimodal task? Subsequently, what are the most suitable multimodal models to capture these interactions? To answer these questions, we propose an information-theoretic approach to quantify the degree of redundancy, uniqueness, and synergy relating input modalities with an output task. We term these three measures as the PID statistics of a multimodal distribution (or PID for short), and introduce two new estimators for these PID statistics that scale to high-dimensional distributions. To validate PID estimation, we conduct extensive experiments on both synthetic datasets where the PID is known and on large-scale multimodal benchmarks where PID estimations are compared with human annotations. Finally, we demonstrate their usefulness in (1) quantifying interactions within multimodal datasets, (2) quantifying interactions captured by multimodal models, (3) principled approaches for model selection, and (4) three real-world case studies engaging with domain experts in pathology, mood prediction, and robotic perception where our framework helps to recommend strong multimodal models for each application.
Synthetic data serves as an alternative in training machine learning models, particularly when real-world data is limited or inaccessible. However, ensuring that synthetic data mirrors the complex nuances of real-world data is a challenging task. This paper addresses this issue by exploring the potential of integrating data-centric AI techniques which profile the data to guide the synthetic data generation process. Moreover, we shed light on the often ignored consequences of neglecting these data profiles during synthetic data generation -- despite seemingly high statistical fidelity. Subsequently, we propose a novel framework to evaluate the integration of data profiles to guide the creation of more representative synthetic data. In an empirical study, we evaluate the performance of five state-of-the-art models for tabular data generation on eleven distinct tabular datasets. The findings offer critical insights into the successes and limitations of current synthetic data generation techniques. Finally, we provide practical recommendations for integrating data-centric insights into the synthetic data generation process, with a specific focus on classification performance, model selection, and feature selection. This study aims to reevaluate conventional approaches to synthetic data generation and promote the application of data-centric AI techniques in improving the quality and effectiveness of synthetic data.
Understanding the similarity of the numerous released large language models (LLMs) has many uses, e.g., simplifying model selection, detecting illegal model reuse, and advancing our understanding of what makes LLMs perform well. In this work, we measure the similarity of representations of a set of LLMs with 7B parameters. Our results suggest that some LLMs are substantially different from others. We identify challenges of using representational similarity measures that suggest the need of careful study of similarity scores to avoid false conclusions.
Diffusion models have made tremendous progress in text-driven image and video generation. Now text-to-image foundation models are widely applied to various downstream image synthesis tasks, such as controllable image generation and image editing, while downstream video synthesis tasks are less explored for several reasons. First, it requires huge memory and computation overhead to train a video generation foundation model. Even with video foundation models, additional costly training is still required for downstream video synthesis tasks. Second, although some works extend image diffusion models into videos in a training-free manner, temporal consistency cannot be well preserved. Finally, these adaption methods are specifically designed for one task and fail to generalize to different tasks. To mitigate these issues, we propose a training-free general-purpose video synthesis framework, coined as {\bf BIVDiff}, via bridging specific image diffusion models and general text-to-video foundation diffusion models. Specifically, we first use a specific image diffusion model (e.g., ControlNet and Instruct Pix2Pix) for frame-wise video generation, then perform Mixed Inversion on the generated video, and finally input the inverted latents into the video diffusion models (e.g., VidRD and ZeroScope) for temporal smoothing. This decoupled framework enables flexible image model selection for different purposes with strong task generalization and high efficiency. To validate the effectiveness and general use of BIVDiff, we perform a wide range of video synthesis tasks, including controllable video generation, video editing, video inpainting, and outpainting.
This paper presents LEMR (Label-Efficient Model Ranking) and introduces the MoraBench Benchmark. LEMR is a novel framework that minimizes the need for costly annotations in model selection by strategically annotating instances from an unlabeled validation set. To evaluate LEMR, we leverage the MoraBench Benchmark, a comprehensive collection of model outputs across diverse scenarios. Our extensive evaluation across 23 different NLP tasks in semi-supervised learning, weak supervision, and prompt selection tasks demonstrates LEMR's effectiveness in significantly reducing labeling costs. Key findings highlight the impact of suitable ensemble methods, uncertainty sampling strategies, and model committee selection in enhancing model ranking accuracy. LEMR, supported by the insights from MoraBench, provides a cost-effective and accurate solution for model selection, especially valuable in resource-constrained environments.
Model selection in the context of bandit optimization is a challenging problem, as it requires balancing exploration and exploitation not only for action selection, but also for model selection. One natural approach is to rely on online learning algorithms that treat different models as experts. Existing methods, however, scale poorly ($\text{poly}M$) with the number of models $M$ in terms of their regret. Our key insight is that, for model selection in linear bandits, we can emulate full-information feedback to the online learner with a favorable bias-variance trade-off. This allows us to develop ALEXP, which has an exponentially improved ($\log M$) dependence on $M$ for its regret. ALEXP has anytime guarantees on its regret, and neither requires knowledge of the horizon $n$, nor relies on an initial purely exploratory stage. Our approach utilizes a novel time-uniform analysis of the Lasso, establishing a new connection between online learning and high-dimensional statistics.
Solving complicated AI tasks with different domains and modalities is a key step toward artificial general intelligence. While there are numerous AI models available for various domains and modalities, they cannot handle complicated AI tasks autonomously. Considering large language models (LLMs) have exhibited exceptional abilities in language understanding, generation, interaction, and reasoning, we advocate that LLMs could act as a controller to manage existing AI models to solve complicated AI tasks, with language serving as a generic interface to empower this. Based on this philosophy, we present HuggingGPT, an LLM-powered agent that leverages LLMs (e.g., ChatGPT) to connect various AI models in machine learning communities (e.g., Hugging Face) to solve AI tasks. Specifically, we use ChatGPT to conduct task planning when receiving a user request, select models according to their function descriptions available in Hugging Face, execute each subtask with the selected AI model, and summarize the response according to the execution results. By leveraging the strong language capability of ChatGPT and abundant AI models in Hugging Face, HuggingGPT can tackle a wide range of sophisticated AI tasks spanning different modalities and domains and achieve impressive results in language, vision, speech, and other challenging tasks, which paves a new way towards the realization of artificial general intelligence.
Studying conditional independence among many variables with few observations is a challenging task. Gaussian Graphical Models (GGMs) tackle this problem by encouraging sparsity in the precision matrix through $l_q$ regularization with $q\leq1$. However, most GMMs rely on the $l_1$ norm because the objective is highly non-convex for sub-$l_1$ pseudo-norms. In the frequentist formulation, the $l_1$ norm relaxation provides the solution path as a function of the shrinkage parameter $\lambda$. In the Bayesian formulation, sparsity is instead encouraged through a Laplace prior, but posterior inference for different $\lambda$ requires repeated runs of expensive Gibbs samplers. Here we propose a general framework for variational inference with matrix-variate Normalizing Flow in GGMs, which unifies the benefits of frequentist and Bayesian frameworks. As a key improvement on previous work, we train with one flow a continuum of sparse regression models jointly for all regularization parameters $\lambda$ and all $l_q$ norms, including non-convex sub-$l_1$ pseudo-norms. Within one model we thus have access to (i) the evolution of the posterior for any $\lambda$ and any $l_q$ (pseudo-) norm, (ii) the marginal log-likelihood for model selection, and (iii) the frequentist solution paths through simulated annealing in the MAP limit.
Pre-trained multi-modal vision-language models (VLMs) are becoming increasingly popular due to their exceptional performance on downstream vision applications, particularly in the few- and zero-shot settings. However, selecting the best-performing VLM for some downstream applications is non-trivial, as it is dataset and task-dependent. Meanwhile, the exhaustive evaluation of all available VLMs on a novel application is not only time and computationally demanding but also necessitates the collection of a labeled dataset for evaluation. As the number of open-source VLM variants increases, there is a need for an efficient model selection strategy that does not require access to a curated evaluation dataset. This paper proposes a novel task and benchmark for efficiently evaluating VLMs' zero-shot performance on downstream applications without access to the downstream task dataset. Specifically, we introduce a new task LOVM: Language-Only Vision Model Selection, where methods are expected to perform both model selection and performance prediction based solely on a text description of the desired downstream application. We then introduced an extensive LOVM benchmark consisting of ground-truth evaluations of 35 pre-trained VLMs and 23 datasets, where methods are expected to rank the pre-trained VLMs and predict their zero-shot performance.
In this work, we study the features extracted by English self-supervised learning (SSL) models in cross-lingual contexts and propose a new metric to predict the quality of feature representations. Using automatic speech recognition (ASR) as a downstream task, we analyze the effect of model size, training objectives, and model architecture on the models' performance as a feature extractor for a set of topologically diverse corpora. We develop a novel metric, the Phonetic-Syntax Ratio (PSR), to measure the phonetic and synthetic information in the extracted representations using deep generalized canonical correlation analysis. Results show the contrastive loss in the wav2vec2.0 objective facilitates more effective cross-lingual feature extraction. There is a positive correlation between PSR scores and ASR performance, suggesting that phonetic information extracted by monolingual SSL models can be used for downstream tasks in cross-lingual settings. The proposed metric is an effective indicator of the quality of the representations and can be useful for model selection.
Big data and large-scale machine learning have had a profound impact on science and engineering, particularly in fields focused on forecasting and prediction. Yet, it is still not clear how we can use the superior pattern matching abilities of machine learning models for scientific discovery. This is because the goals of machine learning and science are generally not aligned. In addition to being accurate, scientific theories must also be causally consistent with the underlying physical process and allow for human analysis, reasoning, and manipulation to advance the field. In this paper, we present a case study on discovering a new symbolic model for oceanic rogue waves from data using causal analysis, deep learning, parsimony-guided model selection, and symbolic regression. We train an artificial neural network on causal features from an extensive dataset of observations from wave buoys, while selecting for predictive performance and causal invariance. We apply symbolic regression to distill this black-box model into a mathematical equation that retains the neural network's predictive capabilities, while allowing for interpretation in the context of existing wave theory. The resulting model reproduces known behavior, generates well-calibrated probabilities, and achieves better predictive scores on unseen data than current theory. This showcases how machine learning can facilitate inductive scientific discovery, and paves the way for more accurate rogue wave forecasting.
Recently, the selection of machine learning model based on only the data distribution without concerning the noise of the data. This study aims to distinguish, which models perform well under noisy data, and establish whether stacking machine learning models actually provide robustness to otherwise weak-to-noise models. The electrochemical data were tested with 12 standalone models and stacking model. This includes XGB, LGBM, RF, GB, ADA, NN, ELAS, LASS, RIDGE, SVM, KNN, DT, and the stacking model. It is found that linear models handle noise well with the average error of (slope) to 1.75 F g-1 up to error per 100% percent noise added; but it suffers from prediction accuracy due to having an average of 60.19 F g-1 estimated at minimal error at 0% noise added. Tree-based models fail in terms of noise handling (average slope is 55.24 F g-1 at 100% percent noise), but it can provide higher prediction accuracy (lowest error of 23.9 F g-1) than that of linear. To address the controversial between prediction accuracy and error handling, the stacking model was constructed, which is not only show high accuracy (intercept of 25.03 F g-1), but it also exhibits good noise handling (slope of 43.58 F g-1), making stacking models a relatively low risk and viable choice for beginner and experienced machine learning research in electrochemistry. Even though neural networks (NN) are gaining popularity in the electrochemistry field. However, this study presents that NN is not suitable for electrochemical data, and improper tuning resulting in a model that is susceptible to noise. Thus, STACK models should provide better benefits in that even with untuned base models, they can achieve an accurate and noise-tolerant model. Overall, this work provides insight into machine learning model selection for electrochemical data, which should aid the understanding of data science in chemistry context.
A mixture of multivariate Poisson-log normal factor analyzers is introduced by imposing constraints on the covariance matrix, which resulted in flexible models for clustering purposes. In particular, a class of eight parsimonious mixture models based on the mixtures of factor analyzers model are introduced. Variational Gaussian approximation is used for parameter estimation, and information criteria are used for model selection. The proposed models are explored in the context of clustering discrete data arising from RNA sequencing studies. Using real and simulated data, the models are shown to give favourable clustering performance. The GitHub R package for this work is available at https://github.com/anjalisilva/mixMPLNFA and is released under the open-source MIT license.
Modern multi-centre randomized controlled trials (MCRCTs) collect massive amounts of tabular data, and are monitored intensively for irregularities by humans. We began by empirically evaluating 6 modern machine learning-based outlier detection algorithms on the task of identifying irregular data in 838 datasets from 7 real-world MCRCTs with a total of 77,001 patients from over 44 countries. Our results reinforce key findings from prior work in the outlier detection literature on data from other domains. Existing algorithms often succeed at identifying irregularities without any supervision, with at least one algorithm exhibiting positive performance 70.6% of the time. However, performance across datasets varies substantially with no single algorithm performing consistently well, motivating new techniques for unsupervised model selection or other means of aggregating potentially discordant predictions from multiple candidate models. We propose the Meta-learned Probabilistic Ensemble (MePE), a simple algorithm for aggregating the predictions of multiple unsupervised models, and show that it performs favourably compared to recent meta-learning approaches for outlier detection model selection. While meta-learning shows promise, small ensembles outperform all forms of meta-learning on average, a negative result that may guide the application of current outlier detection approaches in healthcare and other real-world domains.
This paper introduces a new, unsupervised method for automatic video summarization using ideas from generative adversarial networks but eliminating the discriminator, having a simple loss function, and separating training of different parts of the model. An iterative training strategy is also applied by alternately training the reconstructor and the frame selector for multiple iterations. Furthermore, a trainable mask vector is added to the model in summary generation during training and evaluation. The method also includes an unsupervised model selection algorithm. Results from experiments on two public datasets (SumMe and TVSum) and four datasets we created (Soccer, LoL, MLB, and ShortMLB) demonstrate the effectiveness of each component on the model performance, particularly the iterative training strategy. Evaluations and comparisons with the state-of-the-art methods highlight the advantages of the proposed method in performance, stability, and training efficiency.