LANCE: Exploration and Reflection for LLM-based Textual Attacks on News Recommendation
A reflection-driven and exploration-oriented textual attack framework revealing vulnerabilities in news recommender systems.
EvoWiki: Evaluating LLMs on Evolving Knowledge
A benchmark for evaluating LLMs’ ability to track, update, and reason over evolving knowledge.
Wei Tang
Revisiting Language Models in Neural News Recommender Systems
A systematic evaluation of language-model–based architectures within neural news recommendation pipelines.
Let Me Do It For You: Towards LLM Empowered Recommendation via Tool Learning
ToolRec uses LLM agent and conduct tool-learning to achieve controllable, interpretable, and aligned recommendations.
A+B: A General Generator-Reader Framework for Optimizing LLMs to Unleash Synergy Potential
A generator–reader synergy optimization framework for improving large language models.
Wei Tang
UniRel: Unified Representation and Interaction for Joint Relational Triple Extraction
A unified representation–interaction framework for relational triple extraction.
Wei Tang