PyTorchKR Papers

최신 논문

총 4,899편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2411.06559공식 코드 · Papers with Code

Is Your LLM Secretly a World Model of the Internet? Model-Based Planning for Web Agents

Language agents based on large language models (LLMs) have demonstrated great promise in automating web-based tasks. Recent work has shown that incorporating advanced planning algorithms, e.g., tree search, is advantageous over reactive planning for web agents. However, unlike simulated sandbox environments, real-world environments such as the web are rife with irreversible actions. This undermines the feasibility of backtracking, a cornerstone of (tree) search. Overly relying on test-time search also hurts efficiency. We advocate model-based planning for web agents that employs a world model to simulate and deliberate over the outcome of each candidate action before committing to one. We systematically explore this paradigm by (1) Proposing a model-based planning framework, WebDreamer, which employs LLMs to serve as both world models and value functions; (2) Training specialized LLMs as world models with a scalable data synthesis pipeline. Empirical results demonstrate that WebDreamer achieves substantial performance improvements over reactive baselines. It is competitive, while being 4-5 times more efficient, with tree search in sandbox environments (VisualWebArena) and also works effectively on real-world websites (Online-Mind2Web and Mind2Web-Live). Furthermore, our trained world model, Dreamer-7B, performs comparable to GPT-4o, highlighting the potential of specialized world models for efficient and effective planning in complex web environments.

arXiv:2411.05629

Nowcasting distributions: a functional MIDAS model

We propose a functional MIDAS model to leverage high-frequency information for forecasting and nowcasting distributions observed at a lower frequency. We approximate the low-frequency distribution using Functional Principal Component Analysis and consider a group lasso spike-and-slab prior to identify the relevant predictors in the finite-dimensional SUR-MIDAS approximation of the functional MIDAS model. In our application, we use the model to nowcast the U.S. households' income distribution. Our findings indicate that the model enhances forecast accuracy for the entire target distribution and for key features of the distribution that signal changes in inequality.

arXiv:2411.05470

Model-free portfolio allocation in continuous-time

We present a non-probabilistic, path-by-path framework for studying path-dependent (i.e., where weight is a functional of time and historical time-series), long-only portfolio allocation in continuous-time based on [Chiu & Cont '23], where the fundamental concept of self-financing was introduced, independent of any integration theory. In this article, we extend this concept to a portfolio allocation strategy and characterize it by a path-dependent partial differential equation. We derive the general explicit solution that describes the evolution of wealth in generic markets, including price paths that may not evolve continuously or exhibit variation of any order. Explicit solution examples are provided. As an application of our continuous-time, path-dependent framework, we extend an aggregating algorithm of [Vovk '90] and the universal algorithm of [Cover '91] to continuous-time algorithms that combine multiple strategies into a single strategy. These continuous-time (meta) algorithms take multiple strategies as input (which may themselves be generated by other algorithms) and track the wealth generated by the best individual strategy and the best convex combination of strategies, with tracking error bounds in log wealth of order O(1) and O(ln t), respectively. This work extends Cover's theorem [Cover '91, Thm 6.1] to a continuous-time, model-free setting.

arXiv:2411.04425공식 코드 · Papers with Code

DELIFT: Data Efficient Language model Instruction Fine Tuning

Fine-tuning large language models (LLMs) is essential for enhancing their performance on specific tasks but is often resource-intensive due to redundant or uninformative data. To address this inefficiency, we introduce DELIFT (Data Efficient Language model Instruction Fine-Tuning), a novel algorithm that systematically optimizes data selection across the three key stages of fine-tuning: (1) instruction tuning, (2) task-specific fine-tuning (e.g., reasoning, question-answering), and (3) continual fine-tuning (e.g., incorporating new data versions). Unlike existing methods that focus on single-stage optimization or rely on computationally intensive gradient calculations, DELIFT operates efficiently across all stages. Central to our approach is a pairwise utility metric that quantifies how beneficial a data sample is for improving the model's responses to other samples, effectively measuring the informational value relative to the model's current capabilities. By leveraging different submodular functions applied to this metric, DELIFT selects diverse and optimal subsets that are useful across all stages of fine-tuning. Experiments across various tasks and model scales demonstrate that DELIFT can reduce the fine-tuning data size by up to 70% without compromising performance, offering significant computational savings and outperforming existing methods in both efficiency and efficacy.

arXiv:2411.04876공식 코드 · Papers with Code

Non-Euclidean Mixture Model for Social Network Embedding

It is largely agreed that social network links are formed due to either homophily or social influence. Inspired by this, we aim at understanding the generation of links via providing a novel embedding-based graph formation model. Different from existing graph representation learning, where link generation probabilities are defined as a simple function of the corresponding node embeddings, we model the link generation as a mixture model of the two factors. In addition, we model the homophily factor in spherical space and the influence factor in hyperbolic space to accommodate the fact that (1) homophily results in cycles and (2) influence results in hierarchies in networks. We also design a special projection to align these two spaces. We call this model Non-Euclidean Mixture Model, i.e., NMM. We further integrate NMM with our non-Euclidean graph variational autoencoder (VAE) framework, NMM-GNN. NMM-GNN learns embeddings through a unified framework which uses non-Euclidean GNN encoders, non-Euclidean Gaussian priors, a non-Euclidean decoder, and a novel space unification loss component to unify distinct non-Euclidean geometric spaces. Experiments on public datasets show NMM-GNN significantly outperforms state-of-the-art baselines on social network generation and classification tasks, demonstrating its ability to better explain how the social network is formed.

arXiv:2411.04997공식 코드 · Papers with Code

LLM2CLIP: Powerful Language Model Unlocks Richer Visual Representation

CLIP is a foundational multimodal model that aligns image and text features into a shared representation space via contrastive learning on large-scale image-text pairs. Its effectiveness primarily stems from the use of natural language as rich supervision. Motivated by the remarkable advancements in large language models (LLMs), this work explores how LLMs' superior text understanding and extensive open-world knowledge can enhance CLIP's capability, especially for processing longer and more complex image captions. We propose an efficient post-training strategy that integrates LLMs into pretrained CLIP. To address the challenge posed by the autoregressive nature of LLMs, we introduce a caption-to-caption contrastive fine-tuning framework, significantly enhancing the discriminative quality of LLM outputs. Extensive experiments demonstrate that our approach outperforms LoRA-based methods, achieving nearly fourfold faster training with superior performance. Furthermore, we validate substantial improvements over state-of-the-art models such as CLIP, EVA02, and SigLip2 across various zero-shot multimodal retrieval tasks, cross-lingual retrieval tasks, and multimodal language model pretraining.

arXiv:2411.04975공식 코드 · Papers with Code

SuffixDecoding: Extreme Speculative Decoding for Emerging AI Applications

Speculative decoding is widely adopted to reduce latency in large language model (LLM) inference by leveraging smaller draft models capable of handling diverse user tasks. However, emerging AI applications, such as LLM-based agents, present unique workload characteristics: instead of diverse independent requests, agentic frameworks typically submit repetitive inference requests, such as multi-agent pipelines performing similar subtasks or self-refinement loops iteratively enhancing outputs. These workloads result in long and highly predictable sequences, which current speculative decoding methods do not effectively exploit. To address this gap, we introduce \emph{SuffixDecoding}, a novel method that utilizes efficient suffix trees to cache long token sequences from prompts and previous outputs. By adaptively speculating more tokens when acceptance likelihood is high and fewer when it is low, SuffixDecoding effectively exploits opportunities for longer speculations while conserving computation when those opportunities are limited. Evaluations on agentic benchmarks, including SWE-Bench and Text-to-SQL, demonstrate that SuffixDecoding achieves speedups of up to 5.3$\times$, outperforming state-of-the-art methods -- 2.8$\times$ faster than model-based approaches like EAGLE-2/3 and 1.9$\times$ faster than model-free approaches such as Token Recycling. SuffixDecoding is open-sourced at https://github.com/snowflakedb/ArcticInference.

arXiv:2411.04580

Interpreting the Learned Model in MuZero Planning

MuZero has achieved superhuman performance in various games by using a dynamics network to predict environment dynamics for planning, without relying on simulators. However, the latent states learned by the dynamics network make its planning process opaque. This paper aims to demystify MuZero's model by interpreting the learned latent states. We incorporate observation reconstruction and state consistency into MuZero training and conduct an in-depth analysis to evaluate latent states across two board games: 9x9 Go and Outer-Open Gomoku, and three Atari games: Breakout, Ms. Pacman, and Pong. Our findings reveal that while the dynamics network becomes less accurate over longer simulations, MuZero still performs effectively by using planning to correct errors. Our experiments also show that the dynamics network learns better latent states in board games than in Atari games. These insights contribute to a better understanding of MuZero and offer directions for future research to improve the playing performance, robustness, and interpretability of the MuZero algorithm.

arXiv:2411.04491

Series-to-Series Diffusion Bridge Model

Diffusion models have risen to prominence in time series forecasting, showcasing their robust capability to model complex data distributions. However, their effectiveness in deterministic predictions is often constrained by instability arising from their inherent stochasticity. In this paper, we revisit time series diffusion models and present a comprehensive framework that encompasses most existing diffusion-based methods. Building on this theoretical foundation, we propose a novel diffusion-based time series forecasting model, the Series-to-Series Diffusion Bridge Model ($\mathrm{S^2DBM}$), which leverages the Brownian Bridge process to reduce randomness in reverse estimations and improves accuracy by incorporating informative priors and conditions derived from historical time series data. Experimental results demonstrate that $\mathrm{S^2DBM}$ delivers superior performance in point-to-point forecasting and competes effectively with other diffusion-based models in probabilistic forecasting.

arXiv:2411.00871공식 코드 · Papers with Code

LLaMo: Large Language Model-based Molecular Graph Assistant

Large Language Models (LLMs) have demonstrated remarkable generalization and instruction-following capabilities with instruction tuning. The advancements in LLMs and instruction tuning have led to the development of Large Vision-Language Models (LVLMs). However, the competency of the LLMs and instruction tuning have been less explored in the molecular domain. Thus, we propose LLaMo: Large Language Model-based Molecular graph assistant, which is an end-to-end trained large molecular graph-language model. To bridge the discrepancy between the language and graph modalities, we present the multi-level graph projector that transforms graph representations into graph tokens by abstracting the output representations of each GNN layer and motif representations with the cross-attention mechanism. We also introduce machine-generated molecular graph instruction data to instruction-tune the large molecular graph-language model for general-purpose molecule and language understanding. Our extensive experiments demonstrate that LLaMo shows the best performance on diverse tasks, such as molecular description generation, property prediction, and IUPAC name prediction. The code of LLaMo is available at https://github.com/mlvlab/LLaMo.

arXiv:2411.03288

Model Predictive Control of Collinear Coulomb Spacecraft Formations

A model predictive control scheme to stabilize desired configurations of collinear Coulomb spacecraft formations is derived in this paper. The nonlinearities of the dynamics with respect to the input make this problem difficult to solve, computationally. It is shown that the nonlinearities in the input lead to a finite horizon optimization problem which is a nonconvex quadratically-constrained quadratic program (QCQP). A convex relaxation of the nonconvex QCQP is therefore derived which can be solved quickly using a convex optimization solver. A simulation of a four spacecraft formation is provided which demonstrates why optimizing over a prediction horizon is a prudent approach to Coulomb spacecraft formation control.

arXiv:2411.01908

Frequency-based Design Method for Model-Free Controllers

Model-Free Control (MFC) has been applied to a wide variety of systems in which it has shown its performance. MFC offers "model-free operation", but the controller design requires some information from the nominal plant. This paper introduces a new design method for model-free controllers that uses minimal data about the system and retrieves a set of stable controller configurations. This method is specifically developed for first-order model-free controllers, but can be extended to second-order controllers, and it relies in a frequency analysis of the controller and the plant. The main feature of the design method is decoupling the design of the main control parameter alpha from the rest, providing specific values for it. The efficacy of the proposed method will be showcased with some relevant application examples.

arXiv:2411.03324

A Surrogate Model for Quay Crane Scheduling Problem

In ports, a variety of tasks are carried out, and scheduling these tasks is crucial due to its significant impact on productivity, making the generation of precise plans essential. This study proposes a method to solve the Quay Crane Scheduling Problem (QCSP), a representative task scheduling problem in ports known to be NP-Hard, more quickly and accurately. First, the study suggests a method to create more accurate work plans for Quay Cranes (QCs) by learning from actual port data to accurately predict the working speed of QCs. Next, a Surrogate Model is proposed by combining a Machine Learning (ML) model with a Genetic Algorithm (GA), which is widely used to solve complex optimization problems, enabling faster and more precise exploration of solutions. Unlike methods that use fixed-dimensional chromosome encoding, the proposed methodology can provide solutions for encodings of various dimensions. To validate the performance of the newly proposed methodology, comparative experiments were conducted, demonstrating faster search speeds and improved fitness scores. The method proposed in this study can be applied not only to QCSP but also to various NP-Hard problems, and it opens up possibilities for the further development of advanced search algorithms by combining heuristic algorithms with ML models.

arXiv:2411.00046

CurateGPT: A flexible language-model assisted biocuration tool

Effective data-driven biomedical discovery requires data curation: a time-consuming process of finding, organizing, distilling, integrating, interpreting, annotating, and validating diverse information into a structured form suitable for databases and knowledge bases. Accurate and efficient curation of these digital assets is critical to ensuring that they are FAIR, trustworthy, and sustainable. Unfortunately, expert curators face significant time and resource constraints. The rapid pace of new information being published daily is exceeding their capacity for curation. Generative AI, exemplified by instruction-tuned large language models (LLMs), has opened up new possibilities for assisting human-driven curation. The design philosophy of agents combines the emerging abilities of generative AI with more precise methods. A curator's tasks can be aided by agents for performing reasoning, searching ontologies, and integrating knowledge across external sources, all efforts otherwise requiring extensive manual effort. Our LLM-driven annotation tool, CurateGPT, melds the power of generative AI together with trusted knowledge bases and literature sources. CurateGPT streamlines the curation process, enhancing collaboration and efficiency in common workflows. Compared to direct interaction with an LLM, CurateGPT's agents enable access to information beyond that in the LLM's training data and they provide direct links to the data supporting each claim. This helps curators, researchers, and engineers scale up curation efforts to keep pace with the ever-increasing volume of scientific data.

arXiv:2411.03678공식 코드 · Papers with Code

Multi-model Ensemble Conformal Prediction in Dynamic Environments

Conformal prediction is an uncertainty quantification method that constructs a prediction set for a previously unseen datum, ensuring the true label is included with a predetermined coverage probability. Adaptive conformal prediction has been developed to address data distribution shifts in dynamic environments. However, the efficiency of prediction sets varies depending on the learning model used. Employing a single fixed model may not consistently offer the best performance in dynamic environments with unknown data distribution shifts. To address this issue, we introduce a novel adaptive conformal prediction framework, where the model used for creating prediction sets is selected on the fly from multiple candidate models. The proposed algorithm is proven to achieve strongly adaptive regret over all intervals while maintaining valid coverage. Experiments on real and synthetic datasets corroborate that the proposed approach consistently yields more efficient prediction sets while maintaining valid coverage, outperforming alternative methods.

arXiv:2411.02745

Foundation AI Model for Medical Image Segmentation

Foundation models refer to artificial intelligence (AI) models that are trained on massive amounts of data and demonstrate broad generalizability across various tasks with high accuracy. These models offer versatile, one-for-many or one-for-all solutions, eliminating the need for developing task-specific AI models. Examples of such foundation models include the Chat Generative Pre-trained Transformer (ChatGPT) and the Segment Anything Model (SAM). These models have been trained on millions to billions of samples and have shown wide-ranging and accurate applications in numerous tasks such as text processing (using ChatGPT) and natural image segmentation (using SAM). In medical image segmentation - finding target regions in medical images - there is a growing need for these one-for-many or one-for-all foundation models. Such models could obviate the need to develop thousands of task-specific AI models, which is currently standard practice in the field. They can also be adapted to tasks with datasets too small for effective training. We discuss two paths to achieve foundation models for medical image segmentation and comment on progress, challenges, and opportunities. One path is to adapt or fine-tune existing models, originally developed for natural images, for use with medical images. The second path entails building models from scratch, exclusively training on medical images.

arXiv:2411.03034

HumanVLM: Foundation for Human-Scene Vision-Language Model

Human-scene vision-language tasks are increasingly prevalent in diverse social applications, yet recent advancements predominantly rely on models specifically tailored to individual tasks. Emerging research indicates that large vision-language models (VLMs) can enhance performance across various downstream vision-language understanding tasks. However, general-domain models often underperform in specialized fields. This study introduces a domain-specific Large Vision-Language Model, Human-Scene Vision-Language Model (HumanVLM), designed to provide a foundation for human-scene Vision-Language tasks. Specifically, (1) we create a large-scale human-scene multimodal image-text dataset (HumanCaption-10M) sourced from the Internet to facilitate domain-specific alignment; (2) develop a captioning approach for human-centered images, capturing human faces, bodies, and backgrounds, and construct a high-quality Human-Scene image-text dataset (HumanCaptionHQ, about 311k pairs) that contain as much detailed information as possible about human; (3) Using HumanCaption-10M and HumanCaptionHQ, we train a HumanVLM. In the experiments, we then evaluate our HumanVLM across varous downstream tasks, where it demonstrates superior overall performance among multimodal models of comparable scale, particularly excelling in human-related tasks and significantly outperforming similar models, including Qwen2VL and ChatGPT-4o. HumanVLM, alongside the data introduced, will stimulate the research in human-around fields.

arXiv:2411.02941

A Mamba Foundation Model for Time Series Forecasting

Time series foundation models have demonstrated strong performance in zero-shot learning, making them well-suited for predicting rapidly evolving patterns in real-world applications where relevant training data are scarce. However, most of these models rely on the Transformer architecture, which incurs quadratic complexity as input length increases. To address this, we introduce TSMamba, a linear-complexity foundation model for time series forecasting built on the Mamba architecture. The model captures temporal dependencies through both forward and backward Mamba encoders, achieving high prediction accuracy. To reduce reliance on large datasets and lower training costs, TSMamba employs a two-stage transfer learning process that leverages pretrained Mamba LLMs, allowing effective time series modeling with a moderate training set. In the first stage, the forward and backward backbones are optimized via patch-wise autoregressive prediction; in the second stage, the model trains a prediction head and refines other components for long-term forecasting. While the backbone assumes channel independence to manage varying channel numbers across datasets, a channel-wise compressed attention module is introduced to capture cross-channel dependencies during fine-tuning on specific multivariate datasets. Experiments show that TSMamba's zero-shot performance is comparable to state-of-the-art time series foundation models, despite using significantly less training data. It also achieves competitive or superior full-shot performance compared to task-specific prediction models. The code will be made publicly available.

arXiv:2411.02674

Wave Network: An Ultra-Small Language Model

We propose an innovative token representation and update method in a new ultra-small language model: the Wave network. Specifically, we use a complex vector to represent each token, encoding both global and local semantics of the input text. A complex vector consists of two components: a magnitude vector representing the global semantics of the input text, and a phase vector capturing the relationships between individual tokens and global semantics. Experiments on the AG News text classification task demonstrate that, when generating complex vectors from randomly initialized token embeddings, our single-layer Wave Network achieves 90.91% accuracy with wave interference and 91.66% with wave modulation - outperforming a single Transformer layer using BERT pre-trained embeddings by 19.23% and 19.98%, respectively, and approaching the accuracy of the pre-trained and fine-tuned BERT base model (94.64%). Additionally, compared to BERT base, the Wave Network reduces video memory usage and training time by 77.34% and 85.62% during wave modulation. In summary, we used a 2.4-million-parameter small language model to achieve accuracy comparable to a 100-million-parameter BERT model in text classification.

arXiv:2411.00273

Efficient Model Compression for Bayesian Neural Networks

Model Compression has drawn much attention within the deep learning community recently. Compressing a dense neural network offers many advantages including lower computation cost, deployability to devices of limited storage and memories, and resistance to adversarial attacks. This may be achieved via weight pruning or fully discarding certain input features. Here we demonstrate a novel strategy to emulate principles of Bayesian model selection in a deep learning setup. Given a fully connected Bayesian neural network with spike-and-slab priors trained via a variational algorithm, we obtain the posterior inclusion probability for every node that typically gets lost. We employ these probabilities for pruning and feature selection on a host of simulated and real-world benchmark data and find evidence of better generalizability of the pruned model in all our experiments.