arXiv subjects

Peng Liu

Publications and source records attributed to Peng Liu.

At least 145 records · Page 8Linked to original sources

arXiv2025

BPO: Revisiting Preference Modeling in Direct Preference Optimization

Direct Preference Optimization (DPO) have emerged as a popular method for aligning Large Language Models (LLMs) with human preferences. While DPO effectively preserves the relative ordering between chosen and rejected responses through pairwise ranking losses, it often neglects absolute reward magnitudes. This oversight can decrease the likelihood of chosen responses and increase the risk of generating out-of-distribution responses, leading to poor performance. We term this issue Degraded Chosen Responses (DCR).To address this issue, we propose Balanced Preference Optimization (BPO), a novel framework that dynamically balances the optimization of chosen and rejected responses through two key components: balanced reward margin and gap adaptor. Unlike previous methods, BPO can fundamentally resolve DPO's DCR issue, without introducing additional constraints to the loss function. Experimental results on multiple mathematical reasoning tasks show that BPO significantly outperforms DPO, improving accuracy by +10.1% with Llama-3.1-8B-Instruct (18.8% to 28.9%) and +11.7% with Qwen2.5-Math-7B (35.0% to 46.7%). It also surpasses DPO variants by +3.6% over IPO (43.1%), +5.0% over SLiC (41.7%), and +3.1% over Cal-DPO (43.6%) on the same model. Remarkably, our algorithm requires only a single line of code modification, making it simple to implement and fully compatible with existing DPO-based frameworks.

cs.CL↗

arXiv2025

Dynamic-I2V: Exploring Image-to-Video Generation Models via Multimodal LLM

Recent advancements in image-to-video (I2V) generation have shown promising performance in conventional scenarios. However, these methods still encounter significant challenges when dealing with complex scenes that require a deep understanding of nuanced motion and intricate object-action relationships. To address these challenges, we present Dynamic-I2V, an innovative framework that integrates Multimodal Large Language Models (MLLMs) to jointly encode visual and textual conditions for a diffusion transformer (DiT) architecture. By leveraging the advanced multimodal understanding capabilities of MLLMs, our model significantly improves motion controllability and temporal coherence in synthesized videos. The inherent multimodality of Dynamic-I2V further enables flexible support for diverse conditional inputs, extending its applicability to various downstream generation tasks. Through systematic analysis, we identify a critical limitation in current I2V benchmarks: a significant bias towards favoring low-dynamic videos, stemming from an inadequate balance between motion complexity and visual quality metrics. To resolve this evaluation gap, we propose DIVE - a novel assessment benchmark specifically designed for comprehensive dynamic quality measurement in I2V generation. In conclusion, extensive quantitative and qualitative experiments confirm that Dynamic-I2V attains state-of-the-art performance in image-to-video generation, particularly revealing significant improvements of 42.5%, 7.9%, and 11.8% in dynamic range, controllability, and quality, respectively, as assessed by the DIVE metric in comparison to existing methods.

cs.CV↗

arXiv2025

Unifying Language Agent Algorithms with Graph-based Orchestration Engine for Reproducible Agent Research

Language agents powered by large language models (LLMs) have demonstrated remarkable capabilities in understanding, reasoning, and executing complex tasks. However, developing robust agents presents significant challenges: substantial engineering overhead, lack of standardized components, and insufficient evaluation frameworks for fair comparison. We introduce Agent Graph-based Orchestration for Reasoning and Assessment (AGORA), a flexible and extensible framework that addresses these challenges through three key contributions: (1) a modular architecture with a graph-based workflow engine, efficient memory management, and clean component abstraction; (2) a comprehensive suite of reusable agent algorithms implementing state-of-the-art reasoning approaches; and (3) a rigorous evaluation framework enabling systematic comparison across multiple dimensions. Through extensive experiments on mathematical reasoning and multimodal tasks, we evaluate various agent algorithms across different LLMs, revealing important insights about their relative strengths and applicability. Our results demonstrate that while sophisticated reasoning approaches can enhance agent capabilities, simpler methods like Chain-of-Thought often exhibit robust performance with significantly lower computational overhead. AGORA not only simplifies language agent development but also establishes a foundation for reproducible agent research through standardized evaluation protocols.

cs.CL↗

arXiv2025

FinRipple: Aligning Large Language Models with Financial Market for Event Ripple Effect Awareness

Financial markets exhibit complex dynamics where localized events trigger ripple effects across entities. Previous event studies, constrained by static single-company analyses and simplistic assumptions, fail to capture these ripple effects. While large language models (LLMs) offer emergent reasoning capabilities, their direct application falters due to structural market unawareness and limited capacity to analyze ripple effects. We propose FinRipple, an elegant framework that empowers LLMs with the ability to analyze ripple effects through financial theory-guided large-scale reinforcement learning. We begin by relaxing the assumptions of previous methods, incorporating a time-varying knowledge graph to accurately represent market structure. By seamlessly integrating classical asset pricing theory, we align the LLM with the market, enabling it to predict ripple effects. To the best of our knowledge, we are the first to provide a standardized definition of ripple effect prediction, a task that is extremely important yet unexplored in the financial domain. Extensive experiments demonstrate that FinRipple provides a promising solution to this task.

cs.SI↗

arXiv2025

Robust Lambda-quantiles and extremal distributions

In this paper, we investigate the robust models for $Λ$-quantiles with partial information regarding the loss distribution, where $Λ$-quantiles extend the classical quantiles by replacing the fixed probability level with a probability/loss function $Λ$. We find that, under some assumptions, the robust $Λ$-quantiles equal the $Λ$-quantiles of the extremal distributions. This finding allows us to obtain the robust $Λ$-quantiles by applying the results of robust quantiles in the literature. Our results are applied to uncertainty sets characterized by three different constraints respectively: moment constraints, probability distance constraints via the Wasserstein metric, and marginal constraints in risk aggregation. We obtain some explicit expressions for robust $Λ$-quantiles by deriving the extremal distributions for each uncertainty set. These results are applied to optimal portfolio selection under model uncertainty.

q-fin.MF↗

arXiv2025

H2VU-Benchmark: A Comprehensive Benchmark for Hierarchical Holistic Video Understanding

With the rapid development of multimodal models, the demand for assessing video understanding capabilities has been steadily increasing. However, existing benchmarks for evaluating video understanding exhibit significant limitations in coverage, task diversity, and scene adaptability. These shortcomings hinder the accurate assessment of models' comprehensive video understanding capabilities. To tackle this challenge, we propose a hierarchical and holistic video understanding (H2VU) benchmark designed to evaluate both general video and online streaming video comprehension. This benchmark contributes three key features: Extended video duration: Spanning videos from brief 3-second clips to comprehensive 1.5-hour recordings, thereby bridging the temporal gaps found in current benchmarks. Comprehensive assessment tasks: Beyond traditional perceptual and reasoning tasks, we have introduced modules for countercommonsense comprehension and trajectory state tracking. These additions test the models' deep understanding capabilities beyond mere prior knowledge. Enriched video data: To keep pace with the rapid evolution of current AI agents, we have expanded first-person streaming video datasets. This expansion allows for the exploration of multimodal models' performance in understanding streaming videos from a first-person perspective. Extensive results from H2VU reveal that existing multimodal large language models (MLLMs) possess substantial potential for improvement in our newly proposed evaluation tasks. We expect that H2VU will facilitate advancements in video understanding research by offering a comprehensive and in-depth analysis of MLLMs.

cs.CV↗

arXiv2025

An accelerated semi-proximal ADMM with applications to multi-block sparse optimization problems

As an extension of the alternating direction method of multipliers (ADMM), the semi-proximal ADMM (sPADMM) has been widely used in various fields due to its flexibility and robustness. In this paper, we first show that the two-block sPADMM algorithm can achieve an $O(1/\sqrt{K})$ non-ergodic convergence rate. Then we propose an accelerated sPADMM (AsPADMM) algorithm by introducing extrapolation techniques and incrementing penalty parameters. The proposed AsPADMM algorithm is proven to converge globally to an optimal solution with a non-ergodic convergence rate of $O(1/K)$. Furthermore, the AsPADMM can be extended and combined with the symmetric Gauss-Seidel decomposition to achieve an accelerated ADMM for multi-block problems. Finally, we apply the proposed AsPADMM to solving the multi-block subproblems in difference-of-convex algorithms for robust low-rank tensor completion problems and mixed sparse optimization problems. The numerical results suggest that the acceleration techniques bring about a notable improvement in the convergence speed.

math.OC↗

arXiv2025

LLM-BSCVM: An LLM-Based Blockchain Smart Contract Vulnerability Management Framework

Smart contracts are a key component of the Web 3.0 ecosystem, widely applied in blockchain services and decentralized applications. However, the automated execution feature of smart contracts makes them vulnerable to potential attacks due to inherent flaws, which can lead to severe security risks and financial losses, even threatening the integrity of the entire decentralized finance system. Currently, research on smart contract vulnerabilities has evolved from traditional program analysis methods to deep learning techniques, with the gradual introduction of Large Language Models. However, existing studies mainly focus on vulnerability detection, lacking systematic cause analysis and Vulnerability Repair. To address this gap, we propose LLM-BSCVM, a Large Language Model-based smart contract vulnerability management framework, designed to provide end-to-end vulnerability detection, analysis, repair, and evaluation capabilities for Web 3.0 ecosystem. LLM-BSCVM combines retrieval-augmented generation technology and multi-agent collaboration, introducing a three-stage method of Decompose-Retrieve-Generate. This approach enables smart contract vulnerability management through the collaborative efforts of six intelligent agents, specifically: vulnerability detection, cause analysis, repair suggestion generation, risk assessment, vulnerability repair, and patch evaluation. Experimental results demonstrate that LLM-BSCVM achieves a vulnerability detection accuracy and F1 score exceeding 91\% on benchmark datasets, comparable to the performance of state-of-the-art (SOTA) methods, while reducing the false positive rate from 7.2\% in SOTA methods to 5.1\%, thus enhancing the reliability of vulnerability management. Furthermore, LLM-BSCVM supports continuous security monitoring and governance of smart contracts through a knowledge base hot-swapping dynamic update mechanism.

cs.CR↗

arXiv2025

Mamba-MOC: A Multicategory Remote Object Counting via State Space Model

Multicategory remote object counting is a fundamental task in computer vision, aimed at accurately estimating the number of objects of various categories in remote images. Existing methods rely on CNNs and Transformers, but CNNs struggle to capture global dependencies, and Transformers are computationally expensive, which limits their effectiveness in remote applications. Recently, Mamba has emerged as a promising solution in the field of computer vision, offering a linear complexity for modeling global dependencies. To this end, we propose Mamba-MOC, a mamba-based network designed for multi-category remote object counting, which represents the first application of Mamba to remote sensing object counting. Specifically, we propose a cross-scale interaction module to facilitate the deep integration of hierarchical features. Then we design a context state space model to capture both global and local contextual information and provide local neighborhood information during the scan process. Experimental results in large-scale realistic scenarios demonstrate that our proposed method achieves state-of-the-art performance compared with some mainstream counting algorithms.

cs.CV↗

arXiv2025

Bridging Theory and Experiment in Materials Discovery: Machine-Learning-Assisted Prediction of Synthesizable Structures

Even though thermodynamic energy-based crystal structure prediction (CSP) has revolutionized materials discovery, the energy-driven CSP approaches often struggle to identify experimentally realizable metastable materials synthesized through kinetically controlled pathways, creating a critical gap between theoretical predictions and experimental synthesis. Here, we propose a synthesizability-driven CSP framework that integrates symmetry-guided structure derivation with a Wyckoff encode-based machine-learning model, allowing for the efficient localization of subspaces likely to yield highly synthesizable structures. Within the identified promising subspaces, a structure-based synthesizability evaluation model, fine-tuned using recently synthesized structures to enhance predictive accuracy, is employed in conjunction with ab initio calculations to systematically identify synthesizable candidates. The framework successfully reproduces 13 experimentally known XSe (X = Sc, Ti, Mn, Fe, Ni, Cu, Zn) structures, demonstrating its effectiveness in predicting synthesizable structures. Notably, 92,310 structures are filtered from the 554,054 candidates predicted by GNoME, exhibiting great potential for promising synthesizability. Additionally, eight thermodynamically favorable Hf-X-O (X = Ti, V, and Mn) structures have been identified, among which three HfV$_2$O$_7$ candidates exhibit high synthesizability, presenting viable candidates for experimental realization and potentially associated with experimentally observed temperature-induced phase transitions. This work establishes a data-driven paradigm for machine-learning-assisted inorganic materials synthesis, highlighting its potential to bridge the gap between computational predictions and experimental realization while unlocking new opportunities for the targeted discovery of novel functional materials.

cond-mat.mtrl-sci↗

arXiv2025

A universal scaling law for active diffusion in complex media

Using granular experiments and computer simulations, we investigate the long-time diffusion of active tracers in a broad class of complex media composed of frozen obstacles of diverse structures. By introducing a dimensionless persistence length $Q = v_d τ_r / d_t$, we propose a modified scaling relation that independently collapses experimental and simulation results across active and passive particles, diverse media, and distinct propulsion mechanisms. Our results reveal a universal active diffusion-structure relation that holds across both equilibrium and nonequilibrium regimes, providing a simple predictive framework for active diffusion in complex environments.

cond-mat.soft↗

arXiv2025

Towards Intelligent Edge Sensing for ISCC Network: Joint Multi-Tier DNN Partitioning and Beamforming Design

The combination of Integrated Sensing and Communication (ISAC) and Mobile Edge Computing (MEC) enables devices to simultaneously sense the environment and offload data to the base stations (BS) for intelligent processing, thereby reducing local computational burdens. However, transmitting raw sensing data from ISAC devices to the BS often incurs substantial fronthaul overhead and latency. This paper investigates a three-tier collaborative inference framework enabled by Integrated Sensing, Communication, and Computing (ISCC), where cloud servers, MEC servers, and ISAC devices cooperatively execute different segments of a pre-trained deep neural network (DNN) for intelligent sensing. By offloading intermediate DNN features, the proposed framework can significantly reduce fronthaul transmission load. Furthermore, multiple-input multiple-output (MIMO) technology is employed to enhance both sensing quality and offloading efficiency. To minimize the overall sensing task inference latency across all ISAC devices, we jointly optimize the DNN partitioning strategy, ISAC beamforming, and computational resource allocation at the MEC servers and devices, subject to sensing beampattern constraints. We also propose an efficient two-layer optimization algorithm. In the inner layer, we derive closed-form solutions for computational resource allocation using the Karush-Kuhn-Tucker conditions. Moreover, we design the ISAC beamforming vectors via an iterative method based on the majorization-minimization and weighted minimum mean square error techniques. In the outer layer, we develop a cross-entropy based probabilistic learning algorithm to determine an optimal DNN partitioning strategy. Simulation results demonstrate that the proposed framework substantially outperforms existing two-tier schemes in inference latency.

eess.SP↗

arXiv2025

Efficient and wavelength-tunable second-harmonic generation towards the green gap

Achieving compact and efficient visible laser sources is crucial for a wide range of applications. However traditional semiconductor laser technology faces difficulties in producing high-brightness green light, leaving a green gap in wavelength coverage. Second-harmonic generation (SHG) offers a promising alternative by converting near-infrared sources to visible wavelengths with high efficiency and spectral purity. Here, we demonstrate efficient and tunable SHG within the green spectrum using a high-Q Si3N4 microresonator. A space-charge grating induced by the photogalvanic effect realizes reconfigurable grating numbers and flexible wavelength tuning. Additionally, grating formation dynamics and competition is observed. These findings underscore the potential of silicon nitride as a robust, integrative platform for on-chip, tunable green light sources.

physics.optics↗

arXiv2025

DR.FIX: Automatically Fixing Data Races at Industry Scale

Data races are a prevalent class of concurrency bugs in shared-memory parallel programs, posing significant challenges to software reliability and reproducibility. While there is an extensive body of research on detecting data races and a wealth of practical detection tools across various programming languages, considerably less effort has been directed toward automatically fixing data races at an industrial scale. In large codebases, data races are continuously introduced and exhibit myriad patterns, making automated fixing particularly challenging. In this paper, we tackle the problem of automatically fixing data races at an industrial scale. We present Dr.Fix, a tool that combines large language models (LLMs) with program analysis to generate fixes for data races in real-world settings, effectively addressing a broad spectrum of racy patterns in complex code contexts. Implemented for Go--the programming language widely used in modern microservice architectures where concurrency is pervasive and data races are common--Dr.Fix seamlessly integrates into existing development workflows. We detail the design of Dr.Fix and examine how individual design choices influence the quality of the fixes produced. Over the past 18 months, Dr.Fix has been integrated into developer workflows at Uber demonstrating its practical utility. During this period, Dr.Fix produced patches for 224 (55%) from a corpus of 404 data races spanning various categories; 193 of these patches (86%) were accepted by more than a hundred developers via code reviews and integrated into the codebase.

cs.DC↗

arXiv2025

VLM-R1: A Stable and Generalizable R1-style Large Vision-Language Model

Recently DeepSeek R1 has shown that reinforcement learning (RL) can substantially improve the reasoning capabilities of Large Language Models (LLMs) through a simple yet effective design. The core of R1 lies in its rule-based reward formulation, which leverages tasks with deterministic ground-truth answers to enable precise and stable reward computation. In the visual domain, we similarly observe that a wide range of visual understanding tasks are inherently equipped with well-defined ground-truth annotations. This property makes them naturally compatible with rule-based reward mechanisms. Motivated by this observation, we investigate the extension of R1-style reinforcement learning to Vision-Language Models (VLMs), aiming to enhance their visual reasoning capabilities. To this end, we develop VLM-R1, a dedicated framework designed to harness RL for improving VLMs' performance on general vision-language tasks. Using this framework, we further explore the feasibility of applying RL to visual domain. Experimental results indicate that the RL-based model not only delivers competitive performance on visual understanding tasks but also surpasses Supervised Fine-Tuning (SFT) in generalization ability. Furthermore, we conduct comprehensive ablation studies that uncover a series of noteworthy insights, including the presence of reward hacking in object detection, the emergence of the "OD aha moment", the impact of training data quality, and the scaling behavior of RL across different model sizes. Through these analyses, we aim to deepen the understanding of how reinforcement learning enhances the capabilities of vision-language models, and we hope our findings and open-source contributions will support continued progress in the vision-language RL community. Our code and model are available at https://github.com/om-ai-lab/VLM-R1

cs.CV↗

arXiv2025

Spontaneous Vectorization in the Einstein-Born-Infeld-Vector Model

We investigate spontaneous vectorization in the Einstein-Born-Infeld-Vector (EBIV) model, where a massless vector field is nonminimally coupled to a nonlinear Born-Infeld (BI) electromagnetic field. This coupling results in an effective mass for the vector field in a Born-Infeld black hole (BIBH) background, triggering tachyonic instability. We numerically construct and analyze such vectorized Born-Infeld black holes (VBIBHs), focusing on their domain of existence, thermodynamic properties, and energy distributions in both Reissner-Nordström (RN)-like and Schwarzschild-like backgrounds. In RN-like BI backgrounds, vectorized solutions emerge from the perturbative instability threshold and persist down to extremality, exhibiting higher entropy and lower free energy compared to their unvectorized counterparts. Conversely, in Schwarzschild-like backgrounds, VBIBHs show bifurcation behavior with two coexisting solution branches, only one of which is thermodynamically favored. We reveal a contrasting energy redistribution pattern between the internal and external fields in the two regimes, governed by the competition between the vector field and the nonlinear BI field. Our findings highlight the rich structure of spontaneous vectorization in nonlinear electrodynamics and provide novel insights into black hole physics beyond linear Maxwell theory.

gr-qc↗

arXiv2025

Screened volume law of Holographic Entanglement Entropy in Holographic Spontaneous Vectorization model

We present a holographic study of spontaneous vectorization in the background of an isotropic asymptotically AdS black brane. By extending spontaneous scalarization to vector fields, we demonstrate how the effective mass of the vector field drives tachyonic instability, leading to a transition from the AdS-RN phase to a vectorized phase. Thermodynamic analysis reveals the critical temperature $ T_c $ and coupling $ α_c $ for this transition, with the vectorized phase exhibiting lower free energy. A central discovery is the emergence of a ``screened volume law'' in the holographic entanglement entropy (HEE), a novel phenomenon where the entanglement entropy scales with the subregion size via a screened entropy density distinct from thermal entropy. This arises from a geometric constraint tied to the vanishing of the Christoffel symbol $ Γ^z{}_{xx} $, which defines an effective boundary outside the horizons. Unlike conventional ``entanglement shadows'' in black hole systems, this surface acts as a boundary for minimal surfaces in a translationally invariant geometry. This screening effect suggests the inability of entanglement measure to fully probe the Hilbert space of this thermal system. Additionally, the HEE in the vectorized phase displays non-monotonic temperature dependence. These results establish spontaneous vectorization as a mechanism for generating novel entanglement structures in holographic systems, with implications for quantum information and critical phenomena in strongly coupled systems.

hep-th↗

arXiv2025

Spontaneous Vectorization in the Einstein-Maxwell-Vector Model

We investigate spontaneous vectorization in the Einstein-Maxwell-Vector (EMV) model, introducing a novel mechanism driven by the interplay between electromagnetic and vector fields. A key innovation in our work is the resolution of an apparent divergence in the vector field near the event horizon, achieved by employing a generalized coordinate transformation. This not only extends the domain of existence for vectorized Reissner-Nordström black holes (VRNBHs), but also refines the theoretical understanding of such solutions. We introduce a new concept of combined charge $\sqrt{\tilde{Q}^2 + \tilde{P}^2}$, which better captures the underlying physics of these black holes and provides a unified framework for analyzing thermodynamics and observable phenomena such as light ring structures. Our findings suggest that VRNBHs exhibit enhanced thermodynamic preference and distinctive light ring properties compared to Reissner-Nordström solutions. Moreover, we demonstrate how this combined charge approach reveals connections to two-charge black hole solutions, offering promising avenues for observational verification within the context of effective field theories.

gr-qc↗