Papers by Yucheng Zhu
R^3AG: Retriever Routing for Retrieval-Augmented Generation (2026.acl-long)
Copied to clipboard
| Challenge: | Retrieval-augmented generation (RAG) is often bottlenecked by the “one-size-fits-all” retrieval paradigm, as different queries exhibit distinct preferences for different retrievers. |
| Approach: | They propose a novel routing framework that explicitly models the dynamic alignment between queries and retriever capabilities and decomposes retriever capability into two learnable dimensions: retrieval quality and generation utility. |
| Outcome: | Experiments on knowledge-intensive tasks show that R3AG outperforms both the best individual retrievers and state-of-the-art static routing methods. |
Market-Bench: Benchmarking Large Language Models on Economic and Trade Competition (2026.acl-long)
Copied to clipboard
| Challenge: | Existing LLM benchmarks focus on semantic complexity or quantitative competition, but rarely both simultaneously under economic scarcity. |
| Approach: | They propose a benchmark that evaluates the capabilities of large language models (LLMs) in economically-relevant tasks through economic and trade competition. |
| Outcome: | The proposed model evaluates the capabilities of large language models in economically-relevant tasks through economic and trade competition. |
Discontinuous Named Entity Recognition as Maximal Clique Discovery (2021.acl-long)
Copied to clipboard
| Challenge: | Existing methods for named entity recognition break the recognition process into several sequential steps. |
| Approach: | They propose a method that breaks the recognition process into several sequential steps . they construct a segment graph for each sentence and a grid tagging scheme to learn it . |
| Outcome: | Experiments show that the proposed method outperforms the state-of-the-art model and achieves 5x speedup over the SOTA model. |
Text Augmented Spatial Aware Zero-shot Referring Image Segmentation (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing zero-shot referring image segmentation methods focus on global-level alignment of image-text pairs, neglecting fine-grained matching between referring sentence and local image regions. |
| Approach: | They propose a zero-shot referring image segmentation task that is training-free . they use a mask proposal network and a text-augmented spatial-correction score . |
| Outcome: | The proposed method outperforms state-of-the-art zero-shot referring image segmentation methods. |
DREAM: Disentangling Risks to Enhance Safety Alignment in Multimodal Large Language Models (2025.naacl-long)
Copied to clipboard
Jianyu Liu, Hangyu Guo, Ranjie Duan, Xingyuan Bu, Yancheng He, Shilong Li, Hui Huang, Jiaheng Liu, Yucheng Wang, Chenchen Jing, Xingwei Qu, Xiao Zhang, Pei Wang, Yanan Wu, Jihao Gu, Yangguang Li, Jianke Zhu
| Challenge: | Multimodal Large Language Models (MLLMs) pose unique safety challenges due to their integration of visual and textual data. |
| Approach: | They propose a method to disentangle risks through step-by-step reasoning within multimodal inputs. |
| Outcome: | The proposed approach improves safety alignment in MLLMs by fine-tuning and iterative Reinforcement Learning from AI feedback. |
DivScore: Zero-Shot Detection of LLM-Generated Text in Specialized Domains (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing zero-shot detectors fail when applied to specialized content due to domain shift . DivScore outperforms state-of-the-art detectors in specialized domains . |
| Approach: | They propose a zero-shot detection framework that uses normalized entropy-based scoring and domain knowledge distillation to identify LLM-generated text in specialized domains. |
| Outcome: | The proposed framework outperforms state-of-the-art detectors on medical and legal datasets with 14.4% higher AUROC and 64.0% higher recall. |
Maximal Clique Based Non-Autoregressive Open Information Extraction (2021.emnlp-main)
Copied to clipboard
| Challenge: | Open Information Extraction (OpenIE) aims to discover textual facts from a given sentence. |
| Approach: | They propose a non-autoregressive framework that generates a fact graph and a graph with an edge linking two nodes that belong to the same fact. |
| Outcome: | The proposed framework outperforms current state-of-the-art methods on two benchmark datasets and significantly outperformed the existing ones. |
TPLinker: Single-stage Joint Extraction of Entities and Relations Through Token Pair Linking (2020.coling-main)
Copied to clipboard
| Challenge: | Existing methods to extract entities and relations from unstructured text are susceptible to cascading errors due to the separation of entity detection and relation classification. |
| Approach: | They propose a one-stage joint extraction model that detects overlapping relations while being immune from exposure bias. |
| Outcome: | The proposed model can identify overlapping relations while being immune from exposure bias. |