arXiv subjects

Hua Ye

Publications and source records attributed to Hua Ye.

6 recordsLinked to original sources

arXiv2026

Seeing through the Conflict: Transparent Knowledge Conflict Handling in Retrieval-Augmented Generation

Large language models (LLMs) equipped with retrieval--the Retrieval-Augmented Generation (RAG) paradigm--should combine their parametric knowledge with external evidence, yet in practice they often hallucinate, over-trust noisy snippets, or ignore vital context. We introduce TCR (Transparent Conflict Resolution), a plug-and-play framework that makes this decision process observable and controllable. TCR (i) disentangles semantic match and factual consistency via dual contrastive encoders, (ii) estimates self-answerability to gauge confidence in internal memory, and (iii) feeds the three scalar signals to the generator through a lightweight soft-prompt with SNR-based weighting. Across seven benchmarks TCR improves conflict detection (+5-18 F1), raises knowledge-gap recovery by +21.4 pp and cuts misleading-context overrides by -29.3 pp, while adding only 0.3% parameters. The signals align with human judgements and expose temporal decision patterns.

cs.AI↗

arXiv2025

Aligning by Misaligning: Boundary-aware Curriculum Learning for Multimodal Alignment

Most multimodal models treat every negative pair alike, ignoring the ambiguous negatives that differ from the positive by only a small detail. We propose Boundary-Aware Curriculum with Local Attention (BACL), a lightweight add-on that turns these borderline cases into a curriculum signal. A Boundary-aware Negative Sampler gradually raises difficulty, while a Contrastive Local Attention loss highlights where the mismatch occurs. The two modules are fully differentiable and work with any off-the-shelf dual encoder. Theory predicts a fast O(1/n) error rate; practice shows up to +32% R@1 over CLIP and new SOTA on four large-scale benchmarks, all without extra labels.

cs.LG↗

arXiv2025

Synergy over Discrepancy: A Partition-Based Approach to Multi-Domain LLM Fine-Tuning

Large language models (LLMs) demonstrate impressive generalization abilities, yet adapting them effectively across multiple heterogeneous domains remains challenging due to inter-domain interference. To overcome this challenge, we propose a partition-based multi-stage fine-tuning framework designed to exploit inter-domain synergies while minimizing negative transfer. Our approach strategically partitions domains into subsets (stages) by balancing domain discrepancy, synergy, and model capacity constraints. We theoretically analyze the proposed framework and derive novel generalization bounds that justify our partitioning strategy. Extensive empirical evaluations on various language understanding tasks show that our method consistently outperforms state-of-the-art baselines.

cs.LG↗

arXiv2024

All Robots in One: A New Standard and Unified Dataset for Versatile, General-Purpose Embodied Agents

Embodied AI is transforming how AI systems interact with the physical world, yet existing datasets are inadequate for developing versatile, general-purpose agents. These limitations include a lack of standardized formats, insufficient data diversity, and inadequate data volume. To address these issues, we introduce ARIO (All Robots In One), a new data standard that enhances existing datasets by offering a unified data format, comprehensive sensory modalities, and a combination of real-world and simulated data. ARIO aims to improve the training of embodied AI agents, increasing their robustness and adaptability across various tasks and environments. Building upon the proposed new standard, we present a large-scale unified ARIO dataset, comprising approximately 3 million episodes collected from 258 series and 321,064 tasks. The ARIO standard and dataset represent a significant step towards bridging the gaps of existing data resources. By providing a cohesive framework for data collection and representation, ARIO paves the way for the development of more powerful and versatile embodied AI agents, capable of navigating and interacting with the physical world in increasingly complex and diverse ways. The project is available on https://imaei.github.io/project_pages/ario/

cs.RO↗

arXiv2016

Belle II early physics program of bottomonium spectroscopy

The Belle II experiment at the SuperKEKB collider is a major upgrade of the KEK "$B$ factory" facility in Tsukuba, Japan. Phase 1 commissioning of the main ring of SuperKEKB has started in February 2016 and first physics data will be recorded in 2017 during the so-called Phase 2 commissioning, when the partial Belle II detector will be operated still without its vertex detector. In 2018, the full Belle II detector will be rolled in and physics run will start. In this proceeding, a possible physics program for this early data run at different center-of-mass energies is described, in particular at the $Υ(3S)$ and $Υ(6S)$ resonances, amongst other energy points.

hep-ex↗

arXiv2016

FPGA Online Tracking Algorithm for the PANDA Straw Tube Tracker

A novel FPGA based online tracking algorithm for helix track reconstruction in a solenoidal field, developed for the PANDA spectrometer, is described. Employing the Straw Tube Tracker detector with 4636 straw tubes, the algorithm includes a complex track finder, and a track fitter. Implemented in VHDL, the algorithm is tested on a Xilinx Virtex-4 FX60 FPGA chip with different types of events, at different event rates. A processing time of 7 $μ$s per event for an average of 6 charged tracks is obtained. The momentum resolution is about 3\% (4\%) for $p_t$ ($p_z$) at 1 GeV/c. Comparing to the algorithm running on a CPU chip (single core Intel Xeon E5520 at 2.26 GHz), an improvement of 3 orders of magnitude in processing time is obtained. The algorithm can handle severe overlapping of events which are typical for interaction rates above 10 MHz.

physics.ins-det↗