Papers by Donghun Lee
Pre-trained Language Models Return Distinguishable Probability Distributions to Unfaithfully Hallucinated Texts (2024.findings-emnlp)
Copied to clipboard
| Challenge: | 88-98% of cases return distinguishable generation probability and uncertainty distributions to unfaithfully hallucinated texts, regardless of their size and structure. |
| Approach: | They examine 24 pre-trained language models on 6 data sets to examine their ability to distinguish unfaithfully hallucinated texts. |
| Outcome: | The proposed training algorithm outperforms baseline models while maintaining sound general text quality measures. |
P5: Plug-and-Play Persona Prompting for Personalized Response Selection (2023.emnlp-main)
Copied to clipboard
| Challenge: | a plug-and-play persona prompting system can be used to generate personalized responses for real applications . a recent study shows that dialog context alone is insufficient for personalized response selection . |
| Approach: | They propose a plug-and-play persona prompting method that can be used in real applications . they show that the method performs well in the zero-shot setting . |
| Outcome: | The proposed method performs well in the zero-shot setting, and can be fine-tuned for even better performance. |
Noun-MWP: Math Word Problems Meet Noun Answers (2022.coling-1)
Copied to clipboard
| Challenge: | Existing MWP solvers can handle Noun-MWPs, but they are not as efficient as other models. |
| Approach: | They propose a method to empower existing MWP solvers to handle Noun-MWPs. |
| Outcome: | The proposed model solves Noun-MWPs significantly better than other models and solves conventional MWP problems as well. |
SentenceLDA: Discriminative and Robust Document Representation with Sentence Level Topic Model (2024.eacl-long)
Copied to clipboard
| Challenge: | SentenceLDA is a sentence-level topic model that can be used to discriminate between different contexts. |
| Approach: | They propose a sentence-level topic model that extends the semantic unit from word to sentence and a corpus-level key opinion mining model that uses a lexical property to discriminate between different contexts. |
| Outcome: | The proposed model returns more discriminative document representation than other topic models while maintaining LDA’s elegant probabilistic interpretability. |