Many researchers believe that ConvNets perform well on small or moderately sized datasets, but are not competitive with Vision Transformers when given access to datasets on the web-scale. We challenge this belief by evaluating a performant ConvNet architecture pre-trained on JFT-4B, a large labelled dataset of images often used for training foundation models. We consider pre-training compute budgets between 0.4k and 110k TPU-v4 core compute hours, and train a series of networks of increasing depth and width from the NFNet model family. We observe a log-log scaling law between held out loss and compute budget. After fine-tuning on ImageNet, NFNets match the reported performance of Vision Transformers with comparable compute budgets. Our strongest fine-tuned model achieves a Top-1 accuracy of 90.4%.
相關內容
Foundation models, such as Large Language Models (LLMs), can respond to a wide range of format-free queries without any task-specific data collection or model training, creating various research and application opportunities for the modeling and operation of large-scale power systems. In this paper, we outline how such large foundation model such as GPT-4 are developed, and discuss how they can be leveraged in challenging power and energy system tasks. We first investigate the potential of existing foundation models by validating their performance on four representative tasks across power system domains, including the optimal power flow (OPF), electric vehicle (EV) scheduling, knowledge retrieval for power engineering technical reports, and situation awareness. Our results indicate strong capabilities of such foundation models on boosting the efficiency and reliability of power system operational pipelines. We also provide suggestions and projections on future deployment of foundation models in power system applications.
Visualization of extremely large datasets in static or dynamic form is a huge challenge because most traditional methods cannot deal with big data problems. A new visualization method for big data is proposed based on Projection Pursuit, Guided Tour and Data Nuggets methods, that will help display interesting hidden structures such as clusters, outliers, and other nonlinear structures in big data. The Guided Tour is a dynamic graphical tool for high-dimensional data combining Projection Pursuit and Grand Tour methods. It displays a dynamic sequence of low-dimensional projections obtained by using Projection Pursuit (PP) index functions to navigate the data space. Different PP indices have been developed to detect interesting structures of multivariate data but there are computational problems for big data using the original guided tour with these indices. A new PP index is developed to be computable for big data, with the help of a data compression method called Data Nuggets that reduces large datasets while maintaining the original data structure. Simulation studies are conducted and a real large dataset is used to illustrate the proposed methodology. Static and dynamic graphical tools for big data can be developed based on the proposed PP index to detect nonlinear structures.
Large language models (LLMs) have the remarkable ability to solve new tasks with just a few examples, but they need access to the right tools. Retrieval Augmented Generation (RAG) addresses this problem by retrieving a list of relevant tools for a given task. However, RAG's tool retrieval step requires all the required information to be explicitly present in the query. This is a limitation, as semantic search, the widely adopted tool retrieval method, can fail when the query is incomplete or lacks context. To address this limitation, we propose Context Tuning for RAG, which employs a smart context retrieval system to fetch relevant information that improves both tool retrieval and plan generation. Our lightweight context retrieval model uses numerical, categorical, and habitual usage signals to retrieve and rank context items. Our empirical results demonstrate that context tuning significantly enhances semantic search, achieving a 3.5-fold and 1.5-fold improvement in Recall@K for context retrieval and tool retrieval tasks respectively, and resulting in an 11.6% increase in LLM-based planner accuracy. Additionally, we show that our proposed lightweight model using Reciprocal Rank Fusion (RRF) with LambdaMART outperforms GPT-4 based retrieval. Moreover, we observe context augmentation at plan generation, even after tool retrieval, reduces hallucination.
While there is an immense literature on Bayesian methods for clustering, the multiview case has received little attention. This problem focuses on obtaining distinct but statistically dependent clusterings in a common set of entities for different data types. For example, clustering patients into subgroups with subgroup membership varying according to the domain of the patient variables. A challenge is how to model the across-view dependence between the partitions of patients into subgroups. The complexities of the partition space make standard methods to model dependence, such as correlation, infeasible. In this article, we propose CLustering with Independence Centering (CLIC), a clustering prior that uses a single parameter to explicitly model dependence between clusterings across views. CLIC is induced by the product centered Dirichlet process (PCDP), a novel hierarchical prior that bridges between independent and equivalent partitions. We show appealing theoretic properties, provide a finite approximation and prove its accuracy, present a marginal Gibbs sampler for posterior computation, and derive closed form expressions for the marginal and joint partition distributions for the CLIC model. On synthetic data and in an application to epidemiology, CLIC accurately characterizes view-specific partitions while providing inference on the dependence level.
MIMO processing enables jammer mitigation through spatial filtering, provided that the receiver knows the spatial signature of the jammer interference. Estimating this signature is easy for barrage jammers that transmit continuously and with static signature, but difficult for more sophisticated jammers: Smart jammers may deliberately suspend transmission when the receiver tries to estimate their spatial signature, they may use time-varying beamforming to continuously change their spatial signature, or they may stay mostly silent and jam only specific instants (e.g., transmission of control signals). To deal with such smart jammers, we propose MASH, the first method that indiscriminately mitigates all types of jammers: Assume that the transmitter and receiver share a common secret. Based on this secret, the transmitter embeds (with a linear time-domain transform) its signal in a secret subspace of a higher-dimensional space. The receiver applies a reciprocal linear transform to the receive signal, which (i) raises the legitimate transmit signal from its secret subspace and (ii) provably transforms any jammer into a barrage jammer, which makes estimation and mitigation via MIMO processing straightforward. We show the efficacy of MASH for data transmission in the massive multi-user MIMO uplink.
In recent years, pruning has emerged as a popular technique to reduce the computational complexity and memory footprint of Convolutional Neural Network (CNN) models. Mutual Information (MI) has been widely used as a criterion for identifying unimportant filters to prune. However, existing methods for MI computation suffer from high computational cost and sensitivity to noise, leading to suboptimal pruning performance. We propose a novel method to improve MI computation for CNN pruning, using the spatial aura entropy. The spatial aura entropy is useful for evaluating the heterogeneity in the distribution of the neural activations over a neighborhood, providing information about local features. Our method effectively improves the MI computation for CNN pruning, leading to more robust and efficient pruning. Experimental results on the CIFAR-10 benchmark dataset demonstrate the superiority of our approach in terms of pruning performance and computational efficiency.
Multicalibration is a notion of fairness for predictors that requires them to provide calibrated predictions across a large set of protected groups. Multicalibration is known to be a distinct goal than loss minimization, even for simple predictors such as linear functions. In this work, we consider the setting where the protected groups can be represented by neural networks of size $k$, and the predictors are neural networks of size $n > k$. We show that minimizing the squared loss over all neural nets of size $n$ implies multicalibration for all but a bounded number of unlucky values of $n$. We also give evidence that our bound on the number of unlucky values is tight, given our proof technique. Previously, results of the flavor that loss minimization yields multicalibration were known only for predictors that were near the ground truth, hence were rather limited in applicability. Unlike these, our results rely on the expressivity of neural nets and utilize the representation of the predictor.
2D-based Industrial Anomaly Detection has been widely discussed, however, multimodal industrial anomaly detection based on 3D point clouds and RGB images still has many untouched fields. Existing multimodal industrial anomaly detection methods directly concatenate the multimodal features, which leads to a strong disturbance between features and harms the detection performance. In this paper, we propose Multi-3D-Memory (M3DM), a novel multimodal anomaly detection method with hybrid fusion scheme: firstly, we design an unsupervised feature fusion with patch-wise contrastive learning to encourage the interaction of different modal features; secondly, we use a decision layer fusion with multiple memory banks to avoid loss of information and additional novelty classifiers to make the final decision. We further propose a point feature alignment operation to better align the point cloud and RGB features. Extensive experiments show that our multimodal industrial anomaly detection model outperforms the state-of-the-art (SOTA) methods on both detection and segmentation precision on MVTec-3D AD dataset. Code is available at //github.com/nomewang/M3DM.
Embedding models for deterministic Knowledge Graphs (KG) have been extensively studied, with the purpose of capturing latent semantic relations between entities and incorporating the structured knowledge into machine learning. However, there are many KGs that model uncertain knowledge, which typically model the inherent uncertainty of relations facts with a confidence score, and embedding such uncertain knowledge represents an unresolved challenge. The capturing of uncertain knowledge will benefit many knowledge-driven applications such as question answering and semantic search by providing more natural characterization of the knowledge. In this paper, we propose a novel uncertain KG embedding model UKGE, which aims to preserve both structural and uncertainty information of relation facts in the embedding space. Unlike previous models that characterize relation facts with binary classification techniques, UKGE learns embeddings according to the confidence scores of uncertain relation facts. To further enhance the precision of UKGE, we also introduce probabilistic soft logic to infer confidence scores for unseen relation facts during training. We propose and evaluate two variants of UKGE based on different learning objectives. Experiments are conducted on three real-world uncertain KGs via three tasks, i.e. confidence prediction, relation fact ranking, and relation fact classification. UKGE shows effectiveness in capturing uncertain knowledge by achieving promising results on these tasks, and consistently outperforms baselines on these tasks.
We investigate a lattice-structured LSTM model for Chinese NER, which encodes a sequence of input characters as well as all potential words that match a lexicon. Compared with character-based methods, our model explicitly leverages word and word sequence information. Compared with word-based methods, lattice LSTM does not suffer from segmentation errors. Gated recurrent cells allow our model to choose the most relevant characters and words from a sentence for better NER results. Experiments on various datasets show that lattice LSTM outperforms both word-based and character-based LSTM baselines, achieving the best results.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191