Papers by Fanyou Wu

3 papers
DeTiME: Diffusion-Enhanced Topic Modeling using Encoder-decoder based LLM (2023.findings-emnlp)

Copied to clipboard

Challenge: Neural Topic Models and Large Language Models (LLMs) primarily use contextual embeddings from LLMs, which are not optimal for clustering or topic generation.
Approach: They propose a framework that leverages Encoder-Decoders to generate highly clusterable embeddings that could generate topics that exhibit enhanced clusterability and enhanced semantic coherence compared to existing methods.
Outcome: The proposed framework is efficient to train and exhibits high adaptability, demonstrating its potential for a wide array of applications.
Synthesizing Conversations from Unlabeled Documents using Automatic Response Segmentation (2024.findings-acl)

Copied to clipboard

Challenge: Several datasets have been developed for building conversational question answering systems.
Approach: They propose a robust dialog synthesising method that learns segmentation instead of using sentence boundaries.
Outcome: The proposed method achieves superior quality when compared to WikiDialog . it also improves performance across OR-QuAC benchmarks .
Disentangling Biased Knowledge from Reasoning in Large Language Models via Machine Unlearning (2025.acl-long)

Copied to clipboard

Challenge: Existing approaches to disentangle biased knowledge from reasoning are sub-optimal . entangled data makes curation difficult, leading to inclusion of sensitive, toxic data.
Approach: They propose a framework that selectively removes biased knowledge while preserving reasoning abilities.
Outcome: The proposed framework improves fairness accuracy by 14.7% and reasoning performance by 62.6% across multiple LLMs.

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