arXiv subjects

Alan Zhao

Publications and source records attributed to Alan Zhao.

27 records · Page 2Linked to original sources

arXiv2026

Micro-Macro Retrieval: Reducing Long-Form Hallucination in Large Language Models

Large Language Models (LLMs) achieve impressive performance across many tasks but remain prone to hallucination, especially in long-form generation where redundant retrieved contexts and lengthy reasoning chains amplify factual errors. Recent studies highlight a critical phenomenon: the closer key information appears to the model outputs, the higher the factual accuracy. However, existing retrieval-augmented language models (RALMs) lack effective mechanisms to ensure this proximity - external evidence is injected into reasoning via multi-turn retrieval, but this cannot ensure key information stays close to the outputs. We propose Micro-Macro Retrieval (M2R), a novel retrieve-while-generate framework to fill this gap. At the macro level, M2R retrieves coarse-grained evidence from external sources; at the micro level, it extracts essential results from a key information repository built during reasoning and reuses them while generating answers. This design directly addresses the key-information-to-output proximity bottleneck, effectively reducing hallucination in long-form tasks. M2R is trained with a curriculum learning-based reinforcement learning strategy using customized rule-based rewards, enabling stable acquisition of retrieval and grounding skills. Extensive experiments across different benchmarks demonstrate the effectiveness of M2R, especially in lengthy-context settings.

cs.CL↗

arXiv2025

Modulation groups

Conjectures of Braverman and Kazhdan, Ngô and Sakellaridis have motivated the development of Schwartz spaces for certain spherical varieties. We prove that under suitable assumptions these Schwartz spaces are naturally a representation of a group that we christen the modulation group. This provides a broad generalization of the defining representation of the metaplectic group. The example of a vector space and the zero locus of a quadric cone in an even number of variables are discussed in detail. In both of these cases the modulation group is closely related to algebraic groups, and we propose a conjectural method of linking modulation groups to ind-algebraic groups in general. At the end of the paper we discuss adelization and the relationship between representations of modulation groups and the Poisson summation conjecture.

math.NT↗

arXiv2025

Ambiguity Awareness Optimization: Towards Semantic Disambiguation for Direct Preference Optimization

Direct Preference Optimization (DPO) is a widely used reinforcement learning from human feedback (RLHF) method across various domains. Recent research has increasingly focused on the role of token importance in improving DPO effectiveness. It is observed that identical or semantically similar content (defined as ambiguous content) frequently appears within the preference pairs. We hypothesize that the presence of ambiguous content during DPO training may introduce ambiguity, thereby limiting further improvements in alignment. Through mathematical analysis and proof-of-concept experiments, we reveal that ambiguous content may potentially introduce ambiguities, thereby degrading performance. To address this issue, we introduce Ambiguity Awareness Optimization (AAO), a simple yet effective approach that automatically re-weights ambiguous content to reduce ambiguities by calculating semantic similarity from preference pairs. Through extensive experiments, we demonstrate that AAO consistently and significantly surpasses state-of-the-art approaches in performance, without markedly increasing response length, across multiple model scales and widely adopted benchmark datasets, including AlpacaEval 2, MT-Bench, and Arena-Hard. Specifically, AAO outperforms DPO by up to 8.9 points on AlpacaEval 2 and achieves an improvement of by up to 15.0 points on Arena-Hard.

cs.CL↗

arXiv2025

3D Dynamic Fluid Assets from Single-View Videos with Generative Gaussian Splatting

While the generation of 3D content from single-view images has been extensively studied, the creation of physically consistent 3D dynamic scenes from videos remains in its early stages. We propose a novel framework leveraging generative 3D Gaussian Splatting (3DGS) models to extract and re-simulate 3D dynamic fluid objects from single-view videos using simulation methods. The fluid geometry represented by 3DGS is initially generated and optimized from single-view images, then denoised, densified, and aligned across frames. We estimate the fluid surface velocity using optical flow, propose a mainstream extraction algorithm to refine it. The 3D volumetric velocity field is then derived from the velocity of the fluid's enclosed surface. The velocity field is therewith converted into a divergence-free, grid-based representation, enabling the optimization of simulation parameters through its differentiability across frames. This process outputs simulation-ready fluid assets with physical dynamics closely matching those observed in the source video. Our approach is applicable to various liquid fluids, including inviscid and viscous types, and allows users to edit the output geometry or extend movement durations seamlessly. This automatic method for creating 3D dynamic fluid assets from single-view videos, easily obtainable from the internet, shows great potential for generating large-scale 3D fluid assets at a low cost.

cs.GR↗

arXiv2025

AIMMerging: Adaptive Iterative Model Merging Using Training Trajectories for Language Model Continual Learning

Continual learning (CL) is essential for deploying large language models (LLMs) in dynamic real-world environments without the need for costly retraining. Recent model merging-based methods have attracted significant attention, but they still struggle to effectively manage the trade-off between learning new knowledge and preventing forgetting, a challenge largely stemming from suboptimal number of merges and merging frequency. In this paper, we introduce Adaptive Iterative Model Merging (AimMerging), a novel CL framework that utilizes learning and forgetting signals from the training trajectory to dynamically monitor the model's training status. Guided by dynamic monitoring, the training trajectory-guided merge controller adaptively determines the timing and frequency of iterative fusion, while the rehearsal-based knowledge fusion module computes the merging weights and executes the fusion. Comprehensive experiments on three CL benchmarks with various model sizes (from 770M to 13B) demonstrate that AimMerging achieves significant performance improvements over existing state-of-the-art methods, with an average relative improvement of 80% and 59% on FWT and BWT, respectively. The source code is provided for reproducibility.

cs.CL↗

arXiv2025

Crab: A Unified Audio-Visual Scene Understanding Model with Explicit Cooperation

In recent years, numerous tasks have been proposed to encourage model to develop specified capability in understanding audio-visual scene, primarily categorized into temporal localization, spatial localization, spatio-temporal reasoning, and pixel-level understanding. Instead, human possesses a unified understanding ability for diversified tasks. Therefore, designing an audio-visual model with general capability to unify these tasks is of great value. However, simply joint training for all tasks can lead to interference due to the heterogeneity of audiovisual data and complex relationship among tasks. We argue that this problem can be solved through explicit cooperation among tasks. To achieve this goal, we propose a unified learning method which achieves explicit inter-task cooperation from both the perspectives of data and model thoroughly. Specifically, considering the labels of existing datasets are simple words, we carefully refine these datasets and construct an Audio-Visual Unified Instruction-tuning dataset with Explicit reasoning process (AV-UIE), which clarifies the cooperative relationship among tasks. Subsequently, to facilitate concrete cooperation in learning stage, an interaction-aware LoRA structure with multiple LoRA heads is designed to learn different aspects of audiovisual data interaction. By unifying the explicit cooperation across the data and model aspect, our method not only surpasses existing unified audio-visual model on multiple tasks, but also outperforms most specialized models for certain tasks. Furthermore, we also visualize the process of explicit cooperation and surprisingly find that each LoRA head has certain audio-visual understanding ability. Code and dataset: https://github.com/GeWu-Lab/Crab

cs.CV↗

arXiv2025

Dragen3D: Multiview Geometry Consistent 3D Gaussian Generation with Drag-Based Control

Single-image 3D generation has emerged as a prominent research topic, playing a vital role in virtual reality, 3D modeling, and digital content creation. However, existing methods face challenges such as a lack of multi-view geometric consistency and limited controllability during the generation process, which significantly restrict their usability. % To tackle these challenges, we introduce Dragen3D, a novel approach that achieves geometrically consistent and controllable 3D generation leveraging 3D Gaussian Splatting (3DGS). We introduce the Anchor-Gaussian Variational Autoencoder (Anchor-GS VAE), which encodes a point cloud and a single image into anchor latents and decode these latents into 3DGS, enabling efficient latent-space generation. To enable multi-view geometry consistent and controllable generation, we propose a Seed-Point-Driven strategy: first generate sparse seed points as a coarse geometry representation, then map them to anchor latents via the Seed-Anchor Mapping Module. Geometric consistency is ensured by the easily learned sparse seed points, and users can intuitively drag the seed points to deform the final 3DGS geometry, with changes propagated through the anchor latents. To the best of our knowledge, we are the first to achieve geometrically controllable 3D Gaussian generation and editing without relying on 2D diffusion priors, delivering comparable 3D generation quality to state-of-the-art methods.

cs.CV↗

arXiv2022

Heights of Rational Points on Mordell Curves

We conjecture a lower bound for the minimal canonical height of non-torsion rational points on a natural density 1 subset of the sextic twist family of Mordell curves. We then establish a lower bound that yields a partial result towards this conjecture.

math.NT↗

arXiv2020

The Frequency of Elliptic Curves Over $\mathbb{Q}[i]$ with Fixed Torsion

Mazur's Theorem states that there are precisely 15 possibilities for the torsion subgroup of an elliptic curve defined over the rational numbers. It was previously shown by Harron and Snowden that the number of isomorphism classes of elliptic curves of height up to $X$ that have a specific torsion subgroup $G$ is on the order of $X^{1/{d(G)}}$, for some positive $d(G)$ depending on $G$. We compute $d(G)$ for these groups over $\mathbb{Q}[i]$. Furthermore, in a collection of recent papers it was proven that there are 9 more possibilities for the torsion subgroup in the base field $\mathbb{Q}[i]$. We compute the value of $d(G)$ for these new groups.

math.NT↗