Papers by Shirui Pan

13 papers
BlindGuard: Safeguarding LLM-based Multi-Agent Systems under Unknown Attacks (2026.acl-long)

Copied to clipboard

Challenge: Existing supervised defense methods rely on labeled malicious agents to train a supervised model of malicious behavior.
Approach: They propose an unsupervised defense method that learns without requiring any attack-specific labels or prior knowledge of malicious behaviors.
Outcome: The proposed method detects diverse attack types across MAS with various communication patterns while maintaining superior generalizability compared to baselines.
Beyond Memorization: A Rigorous Evaluation Framework for Medical Knowledge Editing (2026.eacl-long)

Copied to clipboard

Challenge: Existing knowledge editing methods show promising results on general-domain benchmarks, but their effectiveness in the medical domain remains largely unexplored.
Approach: They propose a framework to evaluate medical knowledge editing using model-generated rationales as editing targets.
Outcome: The proposed method improves editing efficacy and generalization in medical models without full retraining.
Medical Code Assignment with Gated Convolution and Note-Code Interaction (2021.findings-acl)

Copied to clipboard

Challenge: Medical code assignment from clinical text is a longstanding challenge due to lengthy semantic information in medical notes.
Approach: They propose a method to capture the semantic information of medical notes and a note-code interaction to automate medical code assignment from clinical text.
Outcome: The proposed method outperforms state-of-the-art models on real-world clinical datasets and is on par with light-weighted baselines.
SQL-ASTRA: Alleviating Sparse Feedback in Agentic SQL via Column-Set Matching and Trajectory Aggregation (2026.findings-acl)

Copied to clipboard

Challenge: Agentic SQL is a framework for multiturn agent learning, but it is limited to single-turn paradigms.
Approach: They propose a framework that provides a universal two-tiered reward mechanism for credit assignment . they propose 'Aggregated Trajectory Reward' to resolve multi-turn credit assignment.
Outcome: The proposed framework outperforms SOTA Arctic-Text2SQL-R1-7B on BIRD and Spider 2.0 using identical models.
Fire Burns, Sword Cuts: Commonsense Inductive Bias for Exploration in Text-based Games (2022.acl-short)

Copied to clipboard

Challenge: Existing RL agents are far away from solving text-based games due to their combinatorially large action spaces that hinders efficient exploration.
Approach: They propose an exploration technique that injects external commonsense knowledge, via a pretrained language model, into the agent during training when the agent is the most uncertain about its next action.
Outcome: The proposed method exhibits improvement on the collected game scores during the training in four out of nine games from Jericho.
How Far are We from Robust Long Abstractive Summarization? (2022.emnlp-main)

Copied to clipboard

Challenge: Abstractive summarization has made tremendous progress in recent years . however, even under a short document setting, abstractive models often generate summaries that are repetitive, ungrammatical, and factually inconsistent with the source.
Approach: They perform fine-grained human annotations to evaluate long document abstractive summarization systems and develop factual consistency metrics.
Outcome: The proposed model can generate more relevant summaries but not factual ones.
Leveraging Information Bottleneck for Scientific Document Summarization (2021.findings-emnlp)

Copied to clipboard

Challenge: Existing methods to extract salient sentences from document are unsupervised and rely on graph-based methods for sentence ranking.
Approach: They propose an unsupervised extractive approach to document level summarization based on the Information Bottleneck principle.
Outcome: The proposed framework can be extended to a multi-view framework by different signals.
Explainable and Fine-Grained Safeguarding of LLM Multi-Agent Systems via Bi-Level Graph Anomaly Detection (2026.acl-long)

Copied to clipboard

Challenge: Existing graph anomaly detection methods rely on coarse sentence-level information and overlook fine-grained lexical cues, limiting their reliability and real-world applicability.
Approach: They propose an explainable and fine-grained safeguarding framework for detecting malicious agents in multi-agent systems (MAS) to incorporate both coarse and fine lexical information for anomalous agent identification.
Outcome: Extensive experiments across diverse MAS topologies and attack scenarios demonstrate robust detection performance and strong interpretability of XG-Guard.
CIA: Inferring the Communication Topology from LLM-based Multi-Agent Systems (2026.acl-long)

Copied to clipboard

Challenge: LLM-based multi-agent systems (MAS) have demonstrated remarkable capabilities in solving complex tasks.
Approach: They propose a communication inference attack that constructs new adversarial queries to induce intermediate agents’ reasoning outputs and models their semantic correlations through the global bias disentanglement and LLM-guided weak supervision.
Outcome: The proposed attack achieves an average AUC of 0.87 and a peak AUC up to 0.99, revealing the privacy risk in MAS.
Shrinking Embeddings for Hyper-Relational Knowledge Graphs (2023.acl-long)

Copied to clipboard

Challenge: Existing studies have focused on binary relational KGs where each fact is represented by a triple.
Approach: They propose a geometric hyper-relational KG embedding method that explicitly models qualifier monotonicity, qualifier implication, and qualifier mutual exclusion.
Outcome: The proposed method outperforms existing methods on three benchmarks of hyper-relational KGs.
LexGenius: An Expert-Level Benchmark for Large Language Models in Legal General Intelligence (2026.findings-acl)

Copied to clipboard

Challenge: Existing benchmarks for legal general intelligence (GI) are result-oriented and do not evaluate the legal intelligence of large language models (LLMs).
Approach: They propose a Chinese legal benchmark for evaluating legal GI in large language models . they use recent legal cases and exam questions to create multiple-choice questions .
Outcome: The proposed benchmarks lack a systematic evaluation of the legal intelligence of large language models (LLMs) the results show that even the best LLMs lagging behind human legal professionals.
Enhancing Large Language Model for Knowledge Graph Completion via Structure-Aware Alignment-Tuning (2025.emnlp-main)

Copied to clipboard

Challenge: Existing knowledge graph completion methods ignore inconsistent representation spaces between natural language and graph structures, leading to duplicate works and time-consuming processes.
Approach: They propose a framework that enhances LLMs for KGC via structure-aware alignment-tuning to align graph embeddings with the natural language space through multi-task contrastive learning.
Outcome: The proposed framework outperforms state-of-the-art methods on two KGC tasks across four benchmark datasets.
Understanding the Information Propagation Effects of Communication Topologies in LLM-based Multi-Agent Systems (2025.emnlp-main)

Copied to clipboard

Challenge: Empirical studies for communication topology design often overlook why and when sparse and dense topologies help or hinder collaboration.
Approach: They propose a topology design approach that balances error suppression and beneficial information propagation by fusing connectivity patterns from dense and sparse graphs.
Outcome: The proposed topology design achieves superior performance across tasks with sparse and dense graphs.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations