Zirui Song

SOMIA: Disassembling ARM Binaries by Knowledge-Aware Transfer Learning

Ke Zhang, Shuaike Dong, Fan Yang, Yutong Zhou et autres

hk, cn (code pays fourni par la source)

Parametric Memory Decoding for Zero-Shot Routing in LoRA-Based External Parametric Memory

Fengxian Ji, Zhuohan Xie, Jingpu Yang, F Zhang et autres

With the rise of parametric memory, LoRA-based External Parametric Memory (EPM) has emerged as a modular solution, but existing routing methods often introduce additional training, deployment, and maintenance overhead. This raises a natural question: can a LoRA-based EPM bank be routed without …

Parametric Memory Decoding for Zero-Shot Routing in LoRA-Based External Parametric Memory

Fengxian Ji, Zhuohan Xie, Jingpu Yang, F Zhang et autres

With the rise of parametric memory, LoRA-based External Parametric Memory (EPM) has emerged as a modular solution, but existing routing methods often introduce additional training, deployment, and maintenance overhead. This raises a natural question: can a LoRA-based EPM bank be routed without …

ae, cn (code pays fourni par la source)

FlashSign: Pose-Free Guidance for Efficient Sign Language Video Generation

Liuzhou Zhang, Zeyu Zhang, Biao Wu, Luyao Tang et autres

Sign language plays a crucial role in bridging communication gaps between the deaf and hard-of-hearing communities. However, existing sign language video generation models often rely on complex intermediate representations, which limits their flexibility and efficiency. In this work, we propose a novel …

FlashSign: Pose-Free Guidance for Efficient Sign Language Video Generation

Liuzhou Zhang, Zeyu Zhang, Biao Wu, Luyao Tang et autres

Sign language plays a crucial role in bridging communication gaps between the deaf and hard-of-hearing communities. However, existing sign language video generation models often rely on complex intermediate representations, which limits their flexibility and efficiency. In this work, we propose a novel …

ps, us (code pays fourni par la source)

ManipLVM-R1: Reinforcement Learning for Reasoning in Embodied Manipulation with Large Vision-Language Models

Zirui Song, Guangxian Ouyang, Mingzhe Li, Yuheng Ji et autres

Large Vision-Language Models (LVLMs) have recently advanced robotic manipulation by leveraging vision for scene perception and language for instruction following. However, existing methods rely heavily on costly human-annotated training datasets, which limits their generalization and causes them to struggle in out-of-domain (OOD) …

ae, cn, au (code pays fourni par la source)

FineState-Bench: Benchmarking State-Conditioned Grounding for Fine-grained GUI State Setting

Fengxian Ji, Jingpu Yang, Zirui Song, Yuanxi Wang et autres

Word Form Matters: LLMs' Semantic Reconstruction under Typoglycemia

Chenxi Wang, Tianle Gu, Zhongyu Wei, Zirui Song et autres

Human readers can efficiently comprehend scrambled words, a phenomenon known as Typoglycemia, primarily by relying on word form; if word form alone is insufficient, they further utilize contextual cues for interpretation. While advanced large language models (LLMs) exhibit similar abilities, the underlying …

Word Form Matters: LLMs’ Semantic Reconstruction under Typoglycemia

Chenxi Wang, Tianle Gu, Zhongyu Wei, Zirui Song et autres

Human readers can efficiently comprehend scrambled words, a phenomenon known as Typoglycemia, primarily by relying on word form; if word form alone is insufficient, they further utilize contextual cues for interpretation.While advanced large language models (LLMs) exhibit similar abilities, the underlying mechanisms …

ae (code pays fourni par la source)

BenchLMM: Benchmarking Cross-Style Visual Capability of Large Multimodal Models

Rizhao Cai, Zirui Song, Dayan Guan, Zhenhao Chen et autres

sg, au, cn, ae (code pays fourni par la source)

Efficient Reinforcement Learning via Decoupling Exploration and Utilization

Jingpu Yang, Helin Wang, Qirui Zhao, Zhecheng Shi et autres

cn, au (code pays fourni par la source)

Research of Chinese relation extraction based on BERT

Zirui Song, Li Wan

Facing massive text data on the Internet, it is meaningful to extract valuable information for building knowledge graphs and knowledge engines. Relation extraction is a widespread issue in current research. This paper proposes the BERT-BiLSTM-CRF model for Chinese relation extraction. This paper …

cn (code pays fourni par la source)