arXiv subjects

Lei Li

Publications and source records attributed to Lei Li.

At least 379 records · Page 21Linked to original sources

arXiv2024

The free boundary problem of an epidemic model with nonlocal diffusions and nonlocal reactions: spreading-vanishing dichotomy

This paper concerns the free boundary problem of an epidemic model. The spatial movements of the infectious agents and the infective humans are approximated by nonlocal diffusion operators. Especially, both the growth rate of the agents and the infective rate of humans are represented by nonlocal reaction terms. Thus our model has four integral terms which bring some diffculties for the study of the corresponding principal eigenvalue problem. Firstly, using some elementray analysis instead of Krein-Rutman theorem and the variational characteristic, we obtain the existence and asymptotic behaviors of principal eigenvalue. Then a spreading-vanishing dichotomy is proved to hold, and the criteria for spreading and vanishing are derived. Lastly, comparing our results with those in the existing works, we discuss the effect of nonlocal reaction term on spreading and vanishing, finding that the more nonlocal reaction terms a model has, the harder spreading happens.

math.AP↗

arXiv2024

A note on accurate pressure calculations of Coulomb systems with periodic boundary conditions

In this note, we address some issues concerning the accurate pressure calculation of Coulomb systems with periodic boundary conditions. First, we prove that the formulas for the excess part of the pressure with Ewald summation also reduce to the ensemble average of one-third of the ratio between the potential energy and the volume so that the comments on our previous work in a recent paper by [Onegin~\emph{et al}.,~J. Phys. A: Math.~Theor.~57 (2024) 205002] are incorrect. Second, we demonstrate that in charge non-neutral systems, the pressure expression must be corrected to include interactions with the neutralizing background. This addresses the issues about pressure computation in LAMMPS raised in the paper by Onegin {\it et al.}. Numerical experiments are performed to verify that the pressure obtained via Ewald summation with corrected terms agrees with the average pressure using thermodynamics for the non-neutral OCP system, and are independent of the splitting parameter in the Ewald summation.

physics.comp-ph↗

arXiv2024

Evaluating Durability: Benchmark Insights into Multimodal Watermarking

With the development of large models, watermarks are increasingly employed to assert copyright, verify authenticity, or monitor content distribution. As applications become more multimodal, the utility of watermarking techniques becomes even more critical. The effectiveness and reliability of these watermarks largely depend on their robustness to various disturbances. However, the robustness of these watermarks in real-world scenarios, particularly under perturbations and corruption, is not well understood. To highlight the significance of robustness in watermarking techniques, our study evaluated the robustness of watermarked content generated by image and text generation models against common real-world image corruptions and text perturbations. Our results could pave the way for the development of more robust watermarking techniques in the future. Our project website can be found at \url{https://mmwatermark-robustness.github.io/}.

cs.CV↗

arXiv2024

Don't Fine-Tune, Decode: Syntax Error-Free Tool Use via Constrained Decoding

Instruction-tuned large language models (LLMs) excel at many tasks but often fail to use external tools due to complicated and unfamiliar syntax constraints. While extensive fine-tuning and prompting can mitigate the issue, these approaches are expensive and hard to generalize. Furthermore, because syntax constraints are only learned implicitly during fine-tuning, models still make frequent syntax errors. Motivated by the fact that these constraints can be better satisfied explicitly with constrained decoding, we propose TOOLDEC, a decoding algorithm using finite state machines to force LLMs to follow tool syntax. Our experiments show that TOOLDEC eliminates all syntax errors, achieving significantly better performance on various base models and benchmarks. More surprisingly, when applied to generalist out-of-the-box LLMs such as Mistral-Instruct, TOOLDEC improves its accuracy in tool use from the initial 0% to an impressive 52%, matching the performance of specialized fine-tuned models such as ToolLLM.

cs.CL↗

arXiv2024

SSFlowNet: Semi-supervised Scene Flow Estimation On Point Clouds With Pseudo Label

In the domain of supervised scene flow estimation, the process of manual labeling is both time-intensive and financially demanding. This paper introduces SSFlowNet, a semi-supervised approach for scene flow estimation, that utilizes a blend of labeled and unlabeled data, optimizing the balance between the cost of labeling and the precision of model training. SSFlowNet stands out through its innovative use of pseudo-labels, mainly reducing the dependency on extensively labeled datasets while maintaining high model accuracy. The core of our model is its emphasis on the intricate geometric structures of point clouds, both locally and globally, coupled with a novel spatial memory feature. This feature is adept at learning the geometric relationships between points over sequential time frames. By identifying similarities between labeled and unlabeled points, SSFlowNet dynamically constructs a correlation matrix to evaluate scene flow dependencies at individual point level. Furthermore, the integration of a flow consistency module within SSFlowNet enhances its capability to consistently estimate flow, an essential aspect for analyzing dynamic scenes. Empirical results demonstrate that SSFlowNet surpasses existing methods in pseudo-label generation and shows adaptability across varying data volumes. Moreover, our semi-supervised training technique yields promising outcomes even with different smaller ratio labeled data, marking a substantial advancement in the field of scene flow estimation.

cs.CV↗

arXiv2024

How Vocabulary Sharing Facilitates Multilingualism in LLaMA?

Large Language Models (LLMs), often show strong performance on English tasks, while exhibiting limitations on other languages. What is an LLM's multilingual capability when it is trained only on certain languages? The underlying mechanism remains unclear. This study endeavors to examine the multilingual capability of LLMs from the vocabulary sharing perspective by conducting an exhaustive analysis across 101 languages. Through the investigation of the performance gap before and after embedding fine-tuning, we discovered four distinct quadrants. By delving into each quadrant we provide actionable and efficient guidelines for tuning these languages. Extensive experiments reveal that existing LLMs possess multilingual capabilities that surpass our expectations, and we can significantly improve the multilingual performance of LLMs based on these attributes of each quadrant~\footnote{\url{https://github.com/CONE-MT/Vocabulary-Sharing-Facilitates-Multilingualism}.}.

cs.CL↗

arXiv2024

KS-Lottery: Finding Certified Lottery Tickets for Multilingual Language Models

The lottery ticket hypothesis posits the existence of ``winning tickets'' within a randomly initialized neural network. Do winning tickets exist for LLMs in fine-tuning scenarios? How can we find such winning tickets? In this paper, we propose KS-Lottery, a method to identify a small subset of LLM parameters highly effective in multilingual fine-tuning. Our key idea is to use Kolmogorov-Smirnov Test to analyze the distribution shift of parameters before and after fine-tuning. We further theoretically prove that KS-Lottery can find the certified winning tickets in the embedding layer, fine-tuning on the found parameters is guaranteed to perform as well as full fine-tuning. Comparing KS-Lottery with other parameter-efficient tuning algorithms on translation tasks, the experimental results show that KS-Lottery finds a much smaller set of parameters for fine-tuning while achieving the comparable performance as full fine-tuning LLM. Surprisingly, we find that fine-tuning 18 tokens' embedding of LLaMA suffices to reach the fine-tuning translation performance~\footnote{https://github.com/CONE-MT/KS-Lottery.}.

cs.CL↗

arXiv2024

TempCompass: Do Video LLMs Really Understand Videos?

Recently, there is a surge in interest surrounding video large language models (Video LLMs). However, existing benchmarks fail to provide a comprehensive feedback on the temporal perception ability of Video LLMs. On the one hand, most of them are unable to distinguish between different temporal aspects (e.g., speed, direction) and thus cannot reflect the nuanced performance on these specific aspects. On the other hand, they are limited in the diversity of task formats (e.g., only multi-choice QA), which hinders the understanding of how temporal perception performance may vary across different types of tasks. Motivated by these two problems, we propose the \textbf{TempCompass} benchmark, which introduces a diversity of temporal aspects and task formats. To collect high-quality test data, we devise two novel strategies: (1) In video collection, we construct conflicting videos that share the same static content but differ in a specific temporal aspect, which prevents Video LLMs from leveraging single-frame bias or language priors. (2) To collect the task instructions, we propose a paradigm where humans first annotate meta-information for a video and then an LLM generates the instruction. We also design an LLM-based approach to automatically and accurately evaluate the responses from Video LLMs. Based on TempCompass, we comprehensively evaluate 8 state-of-the-art (SOTA) Video LLMs and 3 Image LLMs, and reveal the discerning fact that these models exhibit notably poor temporal perception ability. Our data will be available at https://github.com/llyx97/TempCompass.

cs.CV↗

arXiv2024

Multimodal ArXiv: A Dataset for Improving Scientific Comprehension of Large Vision-Language Models

Large vision-language models (LVLMs) excel across diverse tasks involving concrete images from natural scenes. However, their ability to interpret abstract figures, such as geometry shapes and scientific plots, remains limited due to a scarcity of training datasets in scientific domains. To fill this gap, we introduce Multimodal ArXiv, consisting of ArXivCap and ArXivQA, for enhancing LVLMs scientific comprehension. ArXivCap is a figure-caption dataset comprising 6.4M images and 3.9M captions, sourced from 572K ArXiv papers spanning various scientific domains. Drawing from ArXivCap, we introduce ArXivQA, a question-answering dataset generated by prompting GPT-4V based on scientific figures. ArXivQA greatly enhances open-sourced LVLMs' mathematical reasoning capabilities, achieving a 10.4\% absolute accuracy gain on a multimodal mathematical reasoning benchmark. Furthermore, employing ArXivCap, we devise four vision-to-text tasks for benchmarking LVLMs. Evaluation results with state-of-the-art LVLMs underscore their struggle with the nuanced semantics of academic figures, while domain-specific training yields substantial performance gains. Our error analysis uncovers misinterpretations of visual context, recognition errors, and the production of overly simplified captions by current LVLMs, shedding light on future improvements.

cs.CV↗

arXiv2024

MedFLIP: Medical Vision-and-Language Self-supervised Fast Pre-Training with Masked Autoencoder

Within the domain of medical analysis, extensive research has explored the potential of mutual learning between Masked Autoencoders(MAEs) and multimodal data. However, the impact of MAEs on intermodality remains a key challenge. We introduce MedFLIP, a Fast Language-Image Pre-training method for Medical analysis. We explore MAEs for zero-shot learning with crossed domains, which enhances the model's ability to learn from limited data, a common scenario in medical diagnostics. We verify that masking an image does not affect inter-modal learning. Furthermore, we propose the SVD loss to enhance the representation learning for characteristics of medical images, aiming to improve classification accuracy by leveraging the structural intricacies of such data. Our theory posits that masking encourages semantic preservation, robust feature extraction, regularization, domain adaptation, and invariance learning. Lastly, we validate using language will improve the zero-shot performance for the medical image analysis. MedFLIP's scaling of the masking process marks an advancement in the field, offering a pathway to rapid and precise medical image analysis without the traditional computational bottlenecks. Through experiments and validation, MedFLIP demonstrates efficient performance improvements, helps for future research and application in medical diagnostics.

eess.IV↗

arXiv2024

DeCo: Decoupling Token Compression from Semantic Abstraction in Multimodal Large Language Models

The visual projector, which bridges the vision and language modalities and facilitates cross-modal alignment, serves as a crucial component in MLLMs. However, measuring the effectiveness of projectors in vision-language alignment remains under-explored, which currently can only be inferred from the performance of MLLMs on downstream tasks. Motivated by the problem, this study examines the projector module by interpreting the vision-language semantic flow within MLLMs. Specifically, we trace back the semantic relevance flow from generated language tokens to raw visual encoder patches and the intermediate outputs produced by projectors. Our findings reveal that compressive projectors (e.g., QFormer), abstract visual patches into a limited set of semantic concepts, such as objects or attributes, resulting in a 'double abstraction' phenomenon. This involves a first visual semantic abstraction by the projector referring to pre-defined query tokens, and a second extraction by the LLM based on text instructions. The double abstraction is inefficient in training and will result in cumulative vision semantics deficiency. To mitigate this issue, we propose the key insight of 'Decouple Compression from Abstraction (DeCo), that is compressing the visual token number at the patch level by projectors and allowing the LLM to handle visual semantic abstraction entirely. Consequently, we adopt a simple compressor, i.e., 2D Adaptive Pooling, to downsample visual patches in a parameter-free manner. Empirical evaluation demonstrates that DeCo surpasses traditional compressive projectors regarding both performance and efficiency. It achieves performance gains of 0.9%, 7.1%, and 2.9% across the MLLM Benchmarks, Visual Localization, and Open-ended VQA tasks with fewer trainable parameters and faster convergence speed.

cs.CV↗

arXiv2024

3D Neural Edge Reconstruction

Real-world objects and environments are predominantly composed of edge features, including straight lines and curves. Such edges are crucial elements for various applications, such as CAD modeling, surface meshing, lane mapping, etc. However, existing traditional methods only prioritize lines over curves for simplicity in geometric modeling. To this end, we introduce EMAP, a new method for learning 3D edge representations with a focus on both lines and curves. Our method implicitly encodes 3D edge distance and direction in Unsigned Distance Functions (UDF) from multi-view edge maps. On top of this neural representation, we propose an edge extraction algorithm that robustly abstracts parametric 3D edges from the inferred edge points and their directions. Comprehensive evaluations demonstrate that our method achieves better 3D edge reconstruction on multiple challenging datasets. We further show that our learned UDF field enhances neural surface reconstruction by capturing more details.

cs.CV↗

arXiv2024

MindMerger: Efficient Boosting LLM Reasoning in non-English Languages

Reasoning capabilities are crucial for Large Language Models (LLMs), yet a notable gap exists between English and non-English languages. To bridge this disparity, some works fine-tune LLMs to relearn reasoning capabilities in non-English languages, while others replace non-English inputs with an external model's outputs such as English translation text to circumvent the challenge of LLM understanding non-English. Unfortunately, these methods often underutilize the built-in skilled reasoning and useful language understanding capabilities of LLMs. In order to better utilize the minds of reasoning and language understanding in LLMs, we propose a new method, namely MindMerger, which merges LLMs with the external language understanding capabilities from multilingual models to boost the multilingual reasoning performance. Furthermore, a two-step training scheme is introduced to first train to embeded the external capabilities into LLMs and then train the collaborative utilization of the external capabilities and the built-in capabilities in LLMs. Experiments on three multilingual reasoning datasets and a language understanding dataset demonstrate that MindMerger consistently outperforms all baselines, especially in low-resource languages. Without updating the parameters of LLMs, the average accuracy improved by 6.7% and 8.0% across all languages and low-resource languages on the MGSM dataset, respectively.

cs.CL↗

arXiv2024

Maps preserving the truncation of triple products on Cartan factors

Let $\{C_i\}_{i\in Γ_1},$ and $\{D_j\}_{j\in Γ_2},$ be two families of Cartan factors such that all of them have dimension at least $2$, and consider the atomic JBW$^*$-triples $A=\bigoplus\limits_{i\in Γ_1}^{\ell_{\infty}} C_i$ and $B=\bigoplus\limits_{j\in Γ_2}^{\ell_{\infty}} D_j$. Let $Δ:A \to B$ be a {\rm(}non-necessarily linear nor continuous{\rm)} bijection preserving the truncation of triple products in both directions, that is, $$\begin{aligned} \boxed{a \mbox{ is a truncation of } \{b,c,b\}} \Leftrightarrow \boxed{Δ(a) \mbox{ is a truncation of } \{Δ(b),Δ(c),Δ(b)\}} \end{aligned}$$ Assume additionally that the restriction of $Δ$ to each rank-one Cartan factor in $A$, if any, is a continuous mapping. Then we show that $Δ$ is an isometric real linear triple isomorphism. We also study some general properties of bijections preserving the truncation of triple products in both directions between general JB$^*$-triples.

math.OA↗

arXiv2024

Explainable Semantic Communication for Text Tasks

Task-oriented semantic communication has gained increasing attention due to its ability to reduce the amount of transmitted data without sacrificing task performance. Although some prior efforts have been dedicated to developing semantic communications, the semantics in these works remains to be unexplainable. Challenges related to explainable semantic representation and knowledge-based semantic compression have yet to be explored. In this paper, we propose a triplet-based explainable semantic communication (TESC) scheme for representing text semantics efficiently. Specifically, we develop a semantic extraction method to convert text into triplets while using syntactic dependency analysis to enhance semantic completeness. Then, we design a semantic filtering method to further compress the duplicate and task-irrelevant triplets based on prior knowledge. The filtered triplets are encoded and transmitted to the receiver for completing intelligent tasks. Furthermore, we apply the propsed TESC scheme to two emblematic text tasks: sentiment analysis and question answering, in which the semantic codec is meticulously customized for each task. Experimental results demonstrate that 1) TESC scheme outperforms benchmarks in terms of Top-1 accuracy and transmission efficiency, and 2) TESC scheme enjoys about 150% performance gain compared to the traditional communication method.

eess.SP↗

arXiv2024

Dynamics of an epidemic model with nonlocal di?usion and a free boundary

An epidemic model, where the dispersal is approximated by nonlocal diffusion operator and spatial domain has one ?xed boundary and one free boundary, is considered in this paper. Firstly, using some elementary analysis instead of variational characterization, we show the existence and asymptotic behaviors of the principal eigenvalue of a cooperative system which can be used to characterize more epidemic models, not just ours. Then we study the existence, uniqueness and stability of a related steady state problem. Finally, we obtain a rather complete understanding for long time behaviors, spreading-vanishing dichotomy, criteria for spreading and vanishing, and spreading speed. Particularly, we prove that the asymptotic spreading speed of solution component (u; v) is equal to the spreading speed of free boundary which is ?nite if and only if a threshold condition holds for kernel functions.

math.AP↗

arXiv2024

Vibe-Eval: A hard evaluation suite for measuring progress of multimodal language models

We introduce Vibe-Eval: a new open benchmark and framework for evaluating multimodal chat models. Vibe-Eval consists of 269 visual understanding prompts, including 100 of hard difficulty, complete with gold-standard responses authored by experts. Vibe-Eval is open-ended and challenging with dual objectives: (i) vibe checking multimodal chat models for day-to-day tasks and (ii) rigorously testing and probing the capabilities of present frontier models. Notably, our hard set contains >50% questions that all frontier models answer incorrectly. We explore the nuances of designing, evaluating, and ranking models on ultra challenging prompts. We also discuss trade-offs between human and automatic evaluation, and show that automatic model evaluation using Reka Core roughly correlates to human judgment. We offer free API access for the purpose of lightweight evaluation and plan to conduct formal human evaluations for public models that perform well on the Vibe-Eval's automatic scores. We release the evaluation code and data, see https://github.com/reka-ai/reka-vibe-eval

cs.CL↗

arXiv2024

Dynamics for a diffusive epidemic model with a free boundary: spreading-vanishing dichotomy

This paper involves a diffusive epidemic model whose domain has one free boundary with the Stefan boundary condition, and one fixed boundary subject to the usual homogeneous Dirichlet or Neumann condition. By using the standard upper and lower solutions method and the regularity theory, we first study some related steady state problems which help us obtain the exact longtime behaviors of solution component $(u,v)$. Then we prove there exists the unique classical solution whose longtime behaviors are governed by a spreading-vanishing dichotomy. Lastly, the criteria determining when spreading or vanishing happens are given with respect to the basic reproduction number $\mathcal{R}_0$, the initial habitat $[0,h_0]$, the expanding rates $μ_1$ and $μ_2$ as well as the initial function $(u_0,v_0)$. The criteria reveal the effect of the cooperative behaviors of agents and humans on spreading and vanishing.

math.AP↗