Papers by Zhenghong Hao

2 papers
CR-LLM: A Dataset and Optimization for Concept Reasoning of Large Language Models (2024.findings-acl)

Copied to clipboard

Challenge: Existing concept reasoning related datasets suffer from modeledge leakage and context leakage.
Approach: They propose a concept reasoning for large language models with modeledge leakage prevention and context leakage preventive methods to improve the models' conceptual reasoning abilities.
Outcome: The proposed method significantly improves the existing models and reasoning methods, achieving a 7% increase in accuracy compared to CoT and showing better granularity.
Prompts Can Play Lottery Tickets Well: Achieving Lifelong Information Extraction via Lottery Prompt Tuning (2023.acl-long)

Copied to clipboard

Challenge: Existing research on information extraction tasks focuses on one specific task, but in real-world scenarios, new data of different IE tasks and domains come in a stream over time.
Approach: They propose a parameter- and deployment-efficient prompt tuning method to evaluate the UIE system under a “lifelong learning” setting.
Outcome: The proposed method is able to learn new tasks without forgetting old ones and expand knowledge and functionalities without retraining the whole system.

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