Papers by Jinjun Xiong
LORE: Continual Logit Rewriting Fosters Faithful Generation (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Using aspect-oriented summarization as a case study, we propose **LOgit REwriting**, a new controlled generation paradigm which can be faithful to external knowledge and to the LLM’s intentions. |
| Approach: | They propose a controlled generation paradigm which can be faithful to external knowledge and to the LLM's intentions. |
| Outcome: | The proposed paradigm can be faithful to external knowledge and to the LLM's intentions while balancing that with accuracy. |
AnaDE1.0: A Novel Data Set for Benchmarking Analogy Detection and Extraction (2024.eacl-long)
Copied to clipboard
| Challenge: | Textual analogies that make comparisons between two concepts are often used for explaining complex ideas, creative writing, and scientific discovery. |
| Approach: | They propose a task that includes three synergistic tasks: detecting documents containing analogies, extracting text segments that make up the analogy, and identifying the (source and target) concepts being compared. |
| Outcome: | The proposed task performs well on all sub-tasks and smaller models perform better than non-finetuned ChatGPT, suggesting high task difficulty. |
DEER: Descriptive Knowledge Graph for Explaining Entity Relationships (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing knowledge graphs lack two desired features for modeling entity relationships: openness and informativeness. |
| Approach: | They propose a self-supervised learning method to extract relation descriptions with the analysis of dependency patterns and generate relation descriptions using a transformer-based relation description synthesizing model. |
| Outcome: | The proposed system extracts and generates high-quality relation descriptions without human labeling. |
Understanding Jargon: Combining Extraction and Generation for Definition Modeling (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for generating definitions of words/phrases perform poorly on jargon. |
| Approach: | They propose to combine extraction and generation for jargon definition modeling by extracting definitional information from the Web and incorporating extracted definitional data. |
| Outcome: | The proposed method outperforms state-of-the-art models significantly on jargon definitions. |
Open Relation Modeling: Learning to Define Relations between Entities (2022.findings-acl)
Copied to clipboard
| Challenge: | Existing systems identify related entities but do not provide features for exploring relations between entities. |
| Approach: | They propose to teach machines to generate definition-like relation descriptions by letting them learn from defining entities. |
| Outcome: | The proposed model can generate definition-like relation descriptions that capture the representative characteristics of entities. |
A Word is Worth A Thousand Dollars: Adversarial Attack on Tweets Fools Stock Prediction (2022.naacl-main)
Copied to clipboard
| Challenge: | Existing models are vulnerable to adversarial attacks, but their vulnerability is underexplored. |
| Approach: | They propose to concatenate a perturbed but semantically similar tweet into a model that fools stock prediction models. |
| Outcome: | The proposed method achieves consistent success rates and causes significant monetary loss in trading simulation by simply concatenating a perturbed but semantically similar tweet. |
Can Language Models Be Specific? How? (2023.findings-acl)
Copied to clipboard
| Challenge: | Existing pre-trained language models have a preference for more specific answers . however, there may exist multiple answers for a query, while not all answers are equally specific. |
| Approach: | They propose to build a benchmark for specificity testing by forming masked token prediction tasks with prompts. |
| Outcome: | The proposed methods improve the specificity of pre-trained language models without additional training. |