There is an emerging line of research on multimodal instruction tuning, and a line of benchmarks has been proposed for evaluating these models recently. Instead of evaluating the models directly, in this paper, we try to evaluate the Vision-Language Instruction-Tuning (VLIT) datasets. Also, we seek the way of building a dataset for developing an all-powerful VLIT model, which we believe could also be of utility for establishing a grounded protocol for benchmarking VLIT models. For effective evaluation of VLIT datasets that remains an open question, we propose a tune-cross-evaluation paradigm: tuning on one dataset and evaluating on the others in turn. For each single tune-evaluation experiment set, we define the Meta Quality (MQ) as the mean score obtained by a set of caption metrics including BLEU, METEOR, and ROUGE-L to quantify the quality of a certain dataset or a sample. On this basis, to evaluate the comprehensiveness of a dataset, we develop the Dataset Quality (DQ) covering all tune-evaluation sets. To lay the foundation for building a comprehensive dataset and developing an all-powerful model for practical applications, we define the Sample Quality (SQ) to quantify the all-sided quality of each sample. Extensive experiments validate the rationality of the proposed evaluation paradigm. Based on the holistic evaluation, we build a new dataset, REVO-LION (REfining VisiOn-Language InstructiOn tuNing), by collecting samples with higher SQ from each dataset. Remarkably, even with only half of the complete data, the model trained on REVO-LION can achieve the performance comparable to simply adding all VLIT datasets up. Furthermore, REVO-LION not only facilitates the development of a powerful model but also incorporates an evaluation set, which is designed to serve as a convenient benchmark for future research in the field.
相關內容
Recent advances in operations research and machine learning have revived interest in solving complex real-world, large-size traffic control problems. With the increasing availability of road sensor data, deterministic parametric models have proved inadequate in describing the variability of real-world data, especially in congested area of the density-flow diagram. In this paper we estimate the stochastic density-flow relation introducing a nonparametric method called convex quantile regression. The proposed method does not depend on any prior functional form assumptions, but thanks to the concavity constraints, the estimated function satisfies the theoretical properties of the density-flow curve. The second contribution is to develop the new convex quantile regression with bags (CQRb) approach to facilitate practical implementation of CQR to the real-world data. We illustrate the CQRb estimation process using the road sensor data from Finland in years 2016-2018. Our third contribution is to demonstrate the excellent out-of-sample predictive power of the proposed CQRb method in comparison to the standard parametric deterministic approach.
Over the past decade, a long line of research has investigated the distributed complexity landscape of locally checkable labeling (LCL) problems on bounded-degree graphs, culminating in an almost-complete classification on general graphs and a complete classification on trees. The latter states that, on bounded-degree trees, any LCL problem has deterministic worst-case time complexity $O(1)$, $\Theta(\log^* n)$, $\Theta(\log n)$, or $\Theta(n^{1/k})$ for some positive integer $k$, and all of those complexity classes are nonempty. Moreover, randomness helps only for (some) problems with deterministic worst-case complexity $\Theta(\log n)$, and if randomness helps (asymptotically), then it helps exponentially. In this work, we study how many distributed rounds are needed on average per node in order to solve an LCL problem on trees. We obtain a partial classification of the deterministic node-averaged complexity landscape for LCL problems. As our main result, we show that every problem with worst-case round complexity $O(\log n)$ has deterministic node-averaged complexity $O(\log^* n)$. Then we show how using randomization we can speed this up and show that every problem with worst case round complexity $O(\log n)$ has randomized node-averaged complexity $O(1)$. We further establish bounds on the node-averaged complexity of problems with worst-case complexity $\Theta(n^{1/k})$: we show that all these problems have node-averaged complexity $\widetilde{\Omega}(n^{1 / (2^k - 1)})$, and that this lower bound is tight for some problems. The lower bound holds even for the randomized case and the upper bound is a deterministic algorithm.
Topological data analysis has emerged as a powerful tool for extracting the metric, geometric and topological features underlying the data as a multi-resolution summary statistic, and has found applications in several areas where data arises from complex sources. In this paper, we examine the use of topological summary statistics through the lens of statistical inference. We investigate necessary and sufficient conditions under which \textit{valid statistical inference} is possible using {topological summary statistics}. Additionally, we provide examples of models that demonstrate invariance with respect to topological summaries.
6G promises a paradigm shift in which positioning and sensing are inherently integrated, enhancing not only the communication performance but also enabling location- and context-aware services. Historically, positioning and sensing have been viewed through the lens of cost and performance trade-offs, implying an escalated demand for resources, such as radio, physical, and computational resources, for improved performance. However, 6G goes beyond this traditional perspective to encompass a set of broader values, namely sustainability, inclusiveness, and trustworthiness. From a joint industrial/academic perspective, this paper aims to shed light on these important value indicators and their relationship with the conventional key performance indicators in the context of positioning and sensing.
The "Sum-Over-Paths" formalism is a way to symbolically manipulate linear maps that describe quantum systems, and is a tool that is used in formal verification of such systems. We give here a new set of rewrite rules for the formalism, and show that it is complete for "Toffoli-Hadamard", the simplest approximately universal fragment of quantum mechanics. We show that the rewriting is terminating, but not confluent (which is expected from the universality of the fragment). We do so using the connection between Sum-over-Paths and graphical language ZH-calculus, and also show how the axiomatisation translates into the latter. We provide generalisations of the presented rewrite rules, that can prove useful when trying to reduce terms in practice, and we show how to graphically make sense of these new rules. We show how to enrich the rewrite system to reach completeness for the dyadic fragments of quantum computation, used in particular in the Quantum Fourier Transform, and obtained by adding phase gates with dyadic multiples of $\pi$ to the Toffoli-Hadamard gate-set. Finally, we show how to perform sums and concatenation of arbitrary terms, something which is not native in a system designed for analysing gate-based quantum computation, but necessary when considering Hamiltonian-based quantum computation.
A longstanding challenge for the Machine Learning community is the one of developing models that are capable of processing and learning from very long sequences of data. The outstanding results of Transformers-based networks (e.g., Large Language Models) promotes the idea of parallel attention as the key to succeed in such a challenge, obfuscating the role of classic sequential processing of Recurrent Models. However, in the last few years, researchers who were concerned by the quadratic complexity of self-attention have been proposing a novel wave of neural models, which gets the best from the two worlds, i.e., Transformers and Recurrent Nets. Meanwhile, Deep Space-State Models emerged as robust approaches to function approximation over time, thus opening a new perspective in learning from sequential data, followed by many people in the field and exploited to implement a special class of (linear) Recurrent Neural Networks. This survey is aimed at providing an overview of these trends framed under the unifying umbrella of Recurrence. Moreover, it emphasizes novel research opportunities that become prominent when abandoning the idea of processing long sequences whose length is known-in-advance for the more realistic setting of potentially infinite-length sequences, thus intersecting the field of lifelong-online learning from streamed data.
We conduct a systematic study of the approximation properties of Transformer for sequence modeling with long, sparse and complicated memory. We investigate the mechanisms through which different components of Transformer, such as the dot-product self-attention, positional encoding and feed-forward layer, affect its expressive power, and we study their combined effects through establishing explicit approximation rates. Our study reveals the roles of critical parameters in the Transformer, such as the number of layers and the number of attention heads, and these insights also provide natural suggestions for alternative architectures.
While there is much excitement about the potential of large multimodal models (LMM), a comprehensive evaluation is critical to establish their true capabilities and limitations. In support of this aim, we evaluate two state-of-the-art LMMs, GPT-4V and Gemini, on a new visual question answering dataset sourced from an authentic online question answering community. We conduct fine-grained analysis by generating seven types of metadata for nearly 2,000 visual questions, such as image type and the required image processing capabilities. Our zero-shot performance analysis highlights the types of questions that are most challenging for both models, including questions related to "puzzling" topic, with "Identification" user intention, with "Sheet Music" image type, or labeled as "hard" by GPT-4.
Recently, graph neural networks have been gaining a lot of attention to simulate dynamical systems due to their inductive nature leading to zero-shot generalizability. Similarly, physics-informed inductive biases in deep-learning frameworks have been shown to give superior performance in learning the dynamics of physical systems. There is a growing volume of literature that attempts to combine these two approaches. Here, we evaluate the performance of thirteen different graph neural networks, namely, Hamiltonian and Lagrangian graph neural networks, graph neural ODE, and their variants with explicit constraints and different architectures. We briefly explain the theoretical formulation highlighting the similarities and differences in the inductive biases and graph architecture of these systems. We evaluate these models on spring, pendulum, gravitational, and 3D deformable solid systems to compare the performance in terms of rollout error, conserved quantities such as energy and momentum, and generalizability to unseen system sizes. Our study demonstrates that GNNs with additional inductive biases, such as explicit constraints and decoupling of kinetic and potential energies, exhibit significantly enhanced performance. Further, all the physics-informed GNNs exhibit zero-shot generalizability to system sizes an order of magnitude larger than the training system, thus providing a promising route to simulate large-scale realistic systems.
In pace with developments in the research field of artificial intelligence, knowledge graphs (KGs) have attracted a surge of interest from both academia and industry. As a representation of semantic relations between entities, KGs have proven to be particularly relevant for natural language processing (NLP), experiencing a rapid spread and wide adoption within recent years. Given the increasing amount of research work in this area, several KG-related approaches have been surveyed in the NLP research community. However, a comprehensive study that categorizes established topics and reviews the maturity of individual research streams remains absent to this day. Contributing to closing this gap, we systematically analyzed 507 papers from the literature on KGs in NLP. Our survey encompasses a multifaceted review of tasks, research types, and contributions. As a result, we present a structured overview of the research landscape, provide a taxonomy of tasks, summarize our findings, and highlight directions for future work.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191