PyTorchKR Papers

최신 논문

총 1,461편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2311.04222

A Modulation-Format Dependent Closed-form Expression for the Gaussian Noise Model in the Presence of Raman Amplification

A closed-form expression that estimates the nonlinear interference of arbitrary modulation formats in Raman amplified links is presented. Accounting for any pumping schemes and inter-channel stimulated Raman scattering effect, the formula is applied to an optical bandwidth of 20~THz and validated using numerical simulations.

arXiv:2306.03819공식 코드 · Papers with Code

LEACE: Perfect linear concept erasure in closed form

Concept erasure aims to remove specified features from an embedding. It can improve fairness (e.g. preventing a classifier from using gender or race) and interpretability (e.g. removing a concept to observe changes in model behavior). We introduce LEAst-squares Concept Erasure (LEACE), a closed-form method which provably prevents all linear classifiers from detecting a concept while changing the embedding as little as possible, as measured by a broad class of norms. We apply LEACE to large language models with a novel procedure called "concept scrubbing," which erases target concept information from every layer in the network. We demonstrate our method on two tasks: measuring the reliance of language models on part-of-speech information, and reducing gender bias in BERT embeddings. Code is available at https://github.com/EleutherAI/concept-erasure.

arXiv:2311.17778

Unified Binary and Multiclass Margin-Based Classification

The notion of margin loss has been central to the development and analysis of algorithms for binary classification. To date, however, there remains no consensus as to the analogue of the margin loss for multiclass classification. In this work, we show that a broad range of multiclass loss functions, including many popular ones, can be expressed in the relative margin form, a generalization of the margin form of binary losses. The relative margin form is broadly useful for understanding and analyzing multiclass losses as shown by our prior work (Wang and Scott, 2020, 2021). To further demonstrate the utility of this way of expressing multiclass losses, we use it to extend the seminal result of Bartlett et al. (2006) on classification-calibration of binary margin losses to multiclass. We then analyze the class of Fenchel-Young losses, and expand the set of these losses that are known to be classification-calibrated.

arXiv:2311.17722

SenTest: Evaluating Robustness of Sentence Encoders

Contrastive learning has proven to be an effective method for pre-training models using weakly labeled data in the vision domain. Sentence transformers are the NLP counterparts to this architecture, and have been growing in popularity due to their rich and effective sentence representations. Having effective sentence representations is paramount in multiple tasks, such as information retrieval, retrieval augmented generation (RAG), and sentence comparison. Keeping in mind the deployability factor of transformers, evaluating the robustness of sentence transformers is of utmost importance. This work focuses on evaluating the robustness of the sentence encoders. We employ several adversarial attacks to evaluate its robustness. This system uses character-level attacks in the form of random character substitution, word-level attacks in the form of synonym replacement, and sentence-level attacks in the form of intra-sentence word order shuffling. The results of the experiments strongly undermine the robustness of sentence encoders. The models produce significantly different predictions as well as embeddings on perturbed datasets. The accuracy of the models can fall up to 15 percent on perturbed datasets as compared to unperturbed datasets. Furthermore, the experiments demonstrate that these embeddings does capture the semantic and syntactic structure (sentence order) of sentences. However, existing supervised classification strategies fail to leverage this information, and merely function as n-gram detectors.

arXiv:2311.17335공식 코드 · Papers with Code

Towards Emotion Analysis in Short-form Videos: A Large-Scale Dataset and Baseline

Nowadays, short-form videos (SVs) are essential to web information acquisition and sharing in our daily life. The prevailing use of SVs to spread emotions leads to the necessity of conducting video emotion analysis (VEA) towards SVs. Considering the lack of SVs emotion data, we introduce a large-scale dataset named eMotions, comprising 27,996 videos. Meanwhile, we alleviate the impact of subjectivities on labeling quality by emphasizing better personnel allocations and multi-stage annotations. In addition, we provide the category-balanced and test-oriented variants through targeted data sampling. Some commonly used videos, such as facial expressions, have been well studied. However, it is still challenging to analysis the emotions in SVs. Since the broader content diversity brings more distinct semantic gaps and difficulties in learning emotion-related features, and there exists local biases and collective information gaps caused by the emotion inconsistence under the prevalently audio-visual co-expressions. To tackle these challenges, we present an end-to-end audio-visual baseline AV-CANet which employs the video transformer to better learn semantically relevant representations. We further design the Local-Global Fusion Module to progressively capture the correlations of audio-visual features. The EP-CE Loss is then introduced to guide model optimization. Extensive experimental results on seven datasets demonstrate the effectiveness of AV-CANet, while providing broad insights for future works. Besides, we investigate the key components of AV-CANet by ablation studies. Datasets and code will be fully open soon.

arXiv:2311.17435

MM-Narrator: Narrating Long-form Videos with Multimodal In-Context Learning

We present MM-Narrator, a novel system leveraging GPT-4 with multimodal in-context learning for the generation of audio descriptions (AD). Unlike previous methods that primarily focused on downstream fine-tuning with short video clips, MM-Narrator excels in generating precise audio descriptions for videos of extensive lengths, even beyond hours, in an autoregressive manner. This capability is made possible by the proposed memory-augmented generation process, which effectively utilizes both the short-term textual context and long-term visual memory through an efficient register-and-recall mechanism. These contextual memories compile pertinent past information, including storylines and character identities, ensuring an accurate tracking and depicting of story-coherent and character-centric audio descriptions. Maintaining the training-free design of MM-Narrator, we further propose a complexity-based demonstration selection strategy to largely enhance its multi-step reasoning capability via few-shot multimodal in-context learning (MM-ICL). Experimental results on MAD-eval dataset demonstrate that MM-Narrator consistently outperforms both the existing fine-tuning-based approaches and LLM-based approaches in most scenarios, as measured by standard evaluation metrics. Additionally, we introduce the first segment-based evaluator for recurrent text generation. Empowered by GPT-4, this evaluator comprehensively reasons and marks AD generation performance in various extendable dimensions.

arXiv:2311.12880공식 코드 · Papers with Code

Weak-Form Latent Space Dynamics Identification

Recent work in data-driven modeling has demonstrated that a weak formulation of model equations enhances the noise robustness of a wide range of computational methods. In this paper, we demonstrate the power of the weak form to enhance the LaSDI (Latent Space Dynamics Identification) algorithm, a recently developed data-driven reduced order modeling technique. We introduce a weak form-based version WLaSDI (Weak-form Latent Space Dynamics Identification). WLaSDI first compresses data, then projects onto the test functions and learns the local latent space models. Notably, WLaSDI demonstrates significantly enhanced robustness to noise. With WLaSDI, the local latent space is obtained using weak-form equation learning techniques. Compared to the standard sparse identification of nonlinear dynamics (SINDy) used in LaSDI, the variance reduction of the weak form guarantees a robust and precise latent space recovery, hence allowing for a fast, robust, and accurate simulation. We demonstrate the efficacy of WLaSDI vs. LaSDI on several common benchmark examples including viscid and inviscid Burgers', radial advection, and heat conduction. For instance, in the case of 1D inviscid Burgers' simulations with the addition of up to 100% Gaussian white noise, the relative error remains consistently below 6% for WLaSDI, while it can exceed 10,000% for LaSDI. Similarly, for radial advection simulations, the relative errors stay below 15% for WLaSDI, in stark contrast to the potential errors of up to 10,000% with LaSDI. Moreover, speedups of several orders of magnitude can be obtained with WLaSDI. For example applying WLaSDI to 1D Burgers' yields a 140X speedup compared to the corresponding full order model. Python code to reproduce the results in this work is available at (https://github.com/MathBioCU/PyWSINDy_ODE) and (https://github.com/MathBioCU/PyWLaSDI).

arXiv:2311.12915

Neural-Integrated Meshfree (NIM) Method: A differentiable programming-based hybrid solver for computational mechanics

We present the neural-integrated meshfree (NIM) method, a differentiable programming-based hybrid meshfree approach within the field of computational mechanics. NIM seamlessly integrates traditional physics-based meshfree discretization techniques with deep learning architectures. It employs a hybrid approximation scheme, NeuroPU, to effectively represent the solution by combining continuous DNN representations with partition of unity (PU) basis functions associated with the underlying spatial discretization. This neural-numerical hybridization not only enhances the solution representation through functional space decomposition but also reduces both the size of DNN model and the need for spatial gradient computations based on automatic differentiation, leading to a significant improvement in training efficiency. Under the NIM framework, we propose two truly meshfree solvers: the strong form-based NIM (S-NIM) and the local variational form-based NIM (V-NIM). In the S-NIM solver, the strong-form governing equation is directly considered in the loss function, while the V-NIM solver employs a local Petrov-Galerkin approach that allows the construction of variational residuals based on arbitrary overlapping subdomains. This ensures both the satisfaction of underlying physics and the preservation of meshfree property. We perform extensive numerical experiments on both stationary and transient benchmark problems to assess the effectiveness of the proposed NIM methods in terms of accuracy, scalability, generalizability, and convergence properties. Moreover, comparative analysis with other physics-informed machine learning methods demonstrates that NIM, especially V-NIM, significantly enhances both accuracy and efficiency in end-to-end predictive capabilities.

arXiv:2311.10918

Jenga Stacking Based on 6D Pose Estimation for Architectural Form Finding Process

This paper includes a review of current state of the art 6d pose estimation methods, as well as a discussion of which pose estimation method should be used in two types of architectural design scenarios. Taking the latest pose estimation research Gen6d as an example, we make a qualitative assessment of the current openset methods in terms of application level, prediction speed, resistance to occlusion, accuracy, resistance to environmental interference, etc. In addition, we try to combine 6D pose estimation and building wind environment assessment to create tangible architectural design approach, we discuss the limitations of the method and point out the direction in which 6d pose estimation is eager to progress in this scenario.

arXiv:2311.09383

Long-form Question Answering: An Iterative Planning-Retrieval-Generation Approach

Long-form question answering (LFQA) poses a challenge as it involves generating detailed answers in the form of paragraphs, which go beyond simple yes/no responses or short factual answers. While existing QA models excel in questions with concise answers, LFQA requires handling multiple topics and their intricate relationships, demanding comprehensive explanations. Previous attempts at LFQA focused on generating long-form answers by utilizing relevant contexts from a corpus, relying solely on the question itself. However, they overlooked the possibility that the question alone might not provide sufficient information to identify the relevant contexts. Additionally, generating detailed long-form answers often entails aggregating knowledge from diverse sources. To address these limitations, we propose an LFQA model with iterative Planning, Retrieval, and Generation. This iterative process continues until a complete answer is generated for the given question. From an extensive experiment on both an open domain and a technical domain QA dataset, we find that our model outperforms the state-of-the-art models on various textual and factual metrics for the LFQA task.

arXiv:2311.09816공식 코드 · Papers with Code

Downstream Trade-offs of a Family of Text Watermarks

Watermarking involves implanting an imperceptible signal into generated text that can later be detected via statistical tests. A prominent family of watermarking strategies for LLMs embeds this signal by upsampling a (pseudorandomly-chosen) subset of tokens at every generation step. However, such signals alter the model's output distribution and can have unintended effects on its downstream performance. In this work, we evaluate the performance of LLMs watermarked using three different strategies over a diverse suite of tasks including those cast as k-class classification (CLS), multiple choice question answering (MCQ), short-form generation (e.g., open-ended question answering) and long-form generation (e.g., translation) tasks. We find that watermarks (under realistic hyperparameters) can cause significant drops in LLMs' effective utility across all tasks. We observe drops of 10 to 20% in CLS tasks in the average case, which shoot up to 100% in the worst case. We notice degradations of about 7% in MCQ tasks, 10-15% in short-form generation, and 5-15% in long-form generation tasks. Our findings highlight the trade-offs that users should be cognizant of when using watermarked models.

arXiv:2311.08610

Converting Transformers to Polynomial Form for Secure Inference Over Homomorphic Encryption

Designing privacy-preserving deep learning models is a major challenge within the deep learning community. Homomorphic Encryption (HE) has emerged as one of the most promising approaches in this realm, enabling the decoupling of knowledge between the model owner and the data owner. Despite extensive research and application of this technology, primarily in convolutional neural networks, incorporating HE into transformer models has been challenging because of the difficulties in converting these models into a polynomial form. We break new ground by introducing the first polynomial transformer, providing the first demonstration of secure inference over HE with transformers. This includes a transformer architecture tailored for HE, alongside a novel method for converting operators to their polynomial equivalent. This innovation enables us to perform secure inference on LMs with WikiText-103. It also allows us to perform image classification with CIFAR-100 and Tiny-ImageNet. Our models yield results comparable to traditional methods, bridging the performance gap with transformers of similar scale and underscoring the viability of HE for state-of-the-art applications. Finally, we assess the stability of our models and conduct a series of ablations to quantify the contribution of each model component.

arXiv:2311.06740

Aggregation and Closed-Form Results for Nonhomothetic CES Preferences

We provide four novel results for nonhomothetic Constant Elasticity of Substitution preferences (Hanoch, 1975). First, we derive a closed-form representation of the expenditure function of nonhomothetic CES under relatively flexible distributional assumptions of demand and price distribution parameters. Second, we characterize aggregate demand from heterogeneous households in closed-form, assuming that household total expenditures follow an empirically plausible distribution. Third, we leverage these results to study the Euler equation arising from standard intertemporal consumption-saving problems featuring within-period nonhomothetic CES preferences. Finally, we show that nonhomothetic CES expenditure shares arise as the solution of a discrete-choice logit problem.

arXiv:2311.06632

The Exact Determinant of a Specific Class of Sparse Positive Definite Matrices

For a specific class of sparse Gaussian graphical models, we provide a closed-form solution for the determinant of the covariance matrix. In our framework, the graphical interaction model (i.e., the covariance selection model) is equal to replacement product of $\mathcal{K}_{n}$ and $\mathcal{K}_{n-1}$, where $\mathcal{K}_n$ is the complete graph with $n$ vertices. Our analysis is based on taking the Fourier transform of the local factors of the model, which can be viewed as an application of the Normal Factor Graph Duality Theorem and holographic algorithms. The closed-form expression is obtained by applying the Matrix Determinant Lemma on the transformed graphical model. In this context, we will also define a notion of equivalence between two Gaussian graphical models.

arXiv:2311.04232

Pattern formation in dense populations studied by inference of nonlinear diffusion-reaction mechanisms

Reaction-diffusion systems have been proposed as a model for pattern formation and morphogenesis. The Fickian diffusion typically employed in these constructions model the Brownian motion of particles. The biological and chemical elements that form the basis of this process, like cells and proteins, occupy finite mass and volume and interact during migration. We propose a Reaction-diffusion system with Maxwell-Stefan formulation to construct the diffusive flux. This formulation relies on inter-species force balance and provides a more realistic model for interacting elements. We also present a variational system inference-based technique to extract these models from spatiotemporal data for these processes. We show that the inferred models can capture the characteristics of local Turing instability that instigates the pattern formation process. Moreover, the equilibrium solutions of the inferred models form similar patterns to the observed data.

arXiv:2311.04459공식 코드 · Papers with Code

Improving Pacing in Long-Form Story Planning

Existing LLM-based systems for writing long-form stories or story outlines frequently suffer from unnatural pacing, whether glossing over important events or over-elaborating on insignificant details, resulting in a jarring experience for the reader. We propose a CONCrete Outline ConTrol (CONCOCT) system to improve pacing when automatically generating story outlines. We first train a concreteness evaluator to judge which of two events is more concrete (low-level-detailed). This evaluator can then be used to control pacing in hierarchical outline generation; in this work, we explore a vaguest-first expansion procedure that aims for uniform pacing. We further use the evaluator to filter new outline items based on predicted concreteness. Compared to a baseline hierarchical outline generator, humans judge CONCOCT's pacing to be more consistent over 57% of the time across multiple outline lengths; the gains also translate to downstream stories. All code, data, and models are open-sourced.

arXiv:2311.02440

Factored-form Kalman-like implementations under maximum correntropy criterion

The maximum correntropy criterion (MCC) methodology is recognized to be a robust filtering strategy with respect to outliers and shown to outperform the classical Kalman filter (KF) for estimation accuracy in the presence of non-Gaussian noise. However, the numerical stability of the newly proposed MCC-KF estimators in finite precision arithmetic is seldom addressed. In this paper, a family of {\it factored-form} (square-root) algorithms is derived for the MCC-KF and its improved variant, respectively. The family traditionally consists of three factored-form implementations: (i) Cholesky factorization-based algorithms, (ii) modified Cholesky, i.e. UD-based methods, and (iii) the recently established SVD-based filtering. All these strategies are commonly recognized to enhance the numerical robustness of conventional filtering with respect to roundoff errors and, hence, they are the preferred implementations when solving applications with high reliability requirements. Previously, only Cholesky-based IMCC-KF algorithms have been designed. This paper enriches a factored-form family by introducing the UD- and SVD-based methods as well. A special attention is paid to {\it array} algorithms that are proved to be the most numerically stable and, additionally, suitable for parallel implementations. The theoretical properties are discussed and numerical comparison is presented for determining the most reliable implementations.

arXiv:2311.01307공식 코드 · Papers with Code

The Effect of Scaling, Retrieval Augmentation and Form on the Factual Consistency of Language Models

Large Language Models (LLMs) make natural interfaces to factual knowledge, but their usefulness is limited by their tendency to deliver inconsistent answers to semantically equivalent questions. For example, a model might predict both "Anne Redpath passed away in Edinburgh." and "Anne Redpath's life ended in London." In this work, we identify potential causes of inconsistency and evaluate the effectiveness of two mitigation strategies: up-scaling and augmenting the LM with a retrieval corpus. Our results on the LLaMA and Atlas models show that both strategies reduce inconsistency while retrieval augmentation is considerably more efficient. We further consider and disentangle the consistency contributions of different components of Atlas. For all LMs evaluated we find that syntactical form and other evaluation task artifacts impact consistency. Taken together, our results provide a better understanding of the factors affecting the factual consistency of language models.

arXiv:2311.00975

Autonomous Learning of Generative Models with Chemical Reaction Network Ensembles

Can a micron sized sack of interacting molecules autonomously learn an internal model of a complex and fluctuating environment? We draw insights from control theory, machine learning theory, chemical reaction network theory, and statistical physics to develop a general architecture whereby a broad class of chemical systems can autonomously learn complex distributions. Our construction takes the form of a chemical implementation of machine learning's optimization workhorse: gradient descent on the relative entropy cost function. We show how this method can be applied to optimize any detailed balanced chemical reaction network and that the construction is capable of using hidden units to learn complex distributions. This result is then recast as a form of integral feedback control. Finally, due to our use of an explicit physical model of learning, we are able to derive thermodynamic costs and trade-offs associated to this process.

arXiv:2311.00463공식 코드 · Papers with Code

Robust and Conjugate Gaussian Process Regression

To enable closed form conditioning, a common assumption in Gaussian process (GP) regression is independent and identically distributed Gaussian observation noise. This strong and simplistic assumption is often violated in practice, which leads to unreliable inferences and uncertainty quantification. Unfortunately, existing methods for robustifying GPs break closed-form conditioning, which makes them less attractive to practitioners and significantly more computationally expensive. In this paper, we demonstrate how to perform provably robust and conjugate Gaussian process (RCGP) regression at virtually no additional cost using generalised Bayesian inference. RCGP is particularly versatile as it enables exact conjugate closed form updates in all settings where standard GPs admit them. To demonstrate its strong empirical performance, we deploy RCGP for problems ranging from Bayesian optimisation to sparse variational Gaussian processes.