Zhi Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Zhi Li has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Yin Zhang
Zhejiang University
4 shared publications - Fei Yu
Zhejiang University Hangzhou China
2 shared publications Hao Chen
Zhejiang University Hangzhou China
2 shared publications- Li Chenglin
Zhejiang University Hangzhou China
2 shared publications - Qianglong Chen
Zhejiang University Hangzhou China
2 shared publications - Yicheng Li
Zhejiang University Hangzhou China
2 shared publications
Latest Publications
- Open Access
Optimizing Instruction Synthesis: Effective Exploration of Evolutionary Space with Tree Search
Instruction tuning is a crucial technique for aligning language models with humans' actual goals in the real world. Extensive research has highlighted the quality of instruction data is essential for the success of this alignment. However, creating high-quality data manually is labor-intensive and time-consuming, which leads researchers to explore using LLMs to synthesize data. Recent studies have focused on using a stronger LLM to iteratively enhance existing instruction data, showing promising results. Nevertheless, previous work often lacks control over the evolution direction, resulting in high uncertainty in the data synthesis process and low-quality instructions. In this paper, we introduce a general and scalable framework, IDEA-MCTS (Instruction Data Enhancement using Monte Carlo Tree Search), a scalable framework for efficiently synthesizing instructions. With tree search and evaluation models, it can efficiently guide each instruction to evolve into a high-quality form, aiding in instruction fine-tuning. Experimental results show that IDEA-MCTS significantly enhances the seed instruction data, raising the average evaluation scores of quality, diversity, and complexity from 2.19 to 3.81. Furthermore, in open-domain benchmarks, experimental results show that IDEA-MCTS improves the accuracy of real-world instruction-following skills in LLMs by an average of 5\% in low-resource settings.
Open AccessOptimizing Instruction Synthesis: Effective Exploration of Evolutionary Space with Tree Search
Instruction tuning is a crucial technique for aligning language models with humans' actual goals in the real world. Extensive research has highlighted the quality of instruction data is essential for the success of this alignment. However, creating high-quality data manually is labor-intensive and time-consuming, which leads researchers to explore using LLMs to synthesize data. Recent studies have focused on using a stronger LLM to iteratively enhance existing instruction data, showing promising results. Nevertheless, previous work often lacks control over the evolution direction, resulting in high uncertainty in the data synthesis process and low-quality instructions. In this paper, we introduce a general and scalable framework, IDEA-MCTS (Instruction Data Enhancement using Monte Carlo Tree Search), a scalable framework for efficiently synthesizing instructions. With tree search and evaluation models, it can efficiently guide each instruction to evolve into a high-quality form, aiding in instruction fine-tuning. Experimental results show that IDEA-MCTS significantly enhances the seed instruction data, raising the average evaluation scores of quality, diversity, and complexity from 2.19 to 3.81. Furthermore, in open-domain benchmarks, experimental results show that IDEA-MCTS improves the accuracy of real-world instruction-following skills in LLMs by an average of 5\% in low-resource settings.
- Open Access
Improving Faithfulness of Large Language Models in Summarization via Sliding Generation and Self-Consistency
Despite large language models (LLMs) have demonstrated impressive performance in various tasks, they are still suffering from the factual inconsistency problem called hallucinations. For instance, LLMs occasionally generate content that diverges from source article, and prefer to extract information that appears at the beginning and end of the context, especially in long document summarization. Inspired by these findings, we propose to improve the faithfulness of LLMs in summarization by impelling them to process the entire article more fairly and faithfully. We present a novel summary generation strategy, namely SliSum, which exploits the ideas of sliding windows and self-consistency. Specifically, SliSum divides the source article into overlapping windows, and utilizes LLM to generate local summaries for the content in the windows. Finally, SliSum aggregates all local summaries using clustering and majority voting algorithm to produce more faithful summary of entire article. Extensive experiments demonstrate that SliSum significantly improves the faithfulness of diverse LLMs including LLaMA-2, Claude-2 and GPT-3.5 in both short and long text summarization, while maintaining their fluency and informativeness and without additional fine-tuning and resources. We further conduct qualitative and quantitative studies to investigate why SliSum works and impacts of hyperparameters in SliSum on performance.
Open AccessImproving Faithfulness of Large Language Models in Summarization via Sliding Generation and Self-Consistency
Despite large language models (LLMs) have demonstrated impressive performance in various tasks, they are still suffering from the factual inconsistency problem called hallucinations. For instance, LLMs occasionally generate content that diverges from source article, and prefer to extract information that appears at the beginning and end of the context, especially in long document summarization. Inspired by these findings, we propose to improve the faithfulness of LLMs in summarization by impelling them to process the entire article more fairly and faithfully. We present a novel summary generation strategy, namely SliSum, which exploits the ideas of sliding windows and self-consistency. Specifically, SliSum divides the source article into overlapping windows, and utilizes LLM to generate local summaries for the content in the windows. Finally, SliSum aggregates all local summaries using clustering and majority voting algorithm to produce more faithful summary of entire article. Extensive experiments demonstrate that SliSum significantly improves the faithfulness of diverse LLMs including LLaMA-2, Claude-2 and GPT-3.5 in both short and long text summarization, while maintaining their fluency and informativeness and without additional fine-tuning and resources. We further conduct qualitative and quantitative studies to investigate why SliSum works and impacts of hyperparameters in SliSum on performance.
- Open Access
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 1707-1721
Optimizing Instruction Synthesis: Effective Exploration of Evolutionary Space with Tree Search
Li Chenglin, Qianglong Chen, Zhi Li, FengTao FengTao, Yicheng Li, Hao Chen, Fei Yu, Yin Zhang. Findings of the Association for Computational Linguistics: EMNLP 2024. 2024.
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 1707-1721
Open AccessOptimizing Instruction Synthesis: Effective Exploration of Evolutionary Space with Tree Search
Li Chenglin, Qianglong Chen, Zhi Li, FengTao FengTao, Yicheng Li, Hao Chen, Fei Yu, Yin Zhang. Findings of the Association for Computational Linguistics: EMNLP 2024. 2024.
- Open Access
- 1 January 2023
- Published by Association for Computational Linguistics (ACL)
- p. 262-276
Cultural Concept Adaptation on Multimodal Reasoning
Zhi Li, Yin Zhang. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 2023.
- 1 January 2023
- Published by Association for Computational Linguistics (ACL)
- p. 262-276
Open AccessCultural Concept Adaptation on Multimodal Reasoning
Zhi Li, Yin Zhang. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 2023.