Papers by Abhilash Dowpati

2 papers
Ukrainian Resilience: A Dataset for Detection of Help-Seeking Signals Amidst the Chaos of War (2024.findings-emnlp)

Copied to clipboard

Challenge: a new dataset is being developed to categorize posts that show distress or urgency . the dataset could improve humanitarian efforts, allowing for quicker and more targeted help .
Approach: They propose a dataset that brings together social media posts in the Ukrainian language for the detection of help-seeking posts in times of war.
Outcome: The proposed dataset can be used to improve humanitarian efforts . it can be compared with existing datasets and achieve an accuracy of 81.15% .
French GossipPrompts: Dataset For Prevention of Generating French Gossip Stories By LLMs (2024.eacl-short)

Copied to clipboard

Challenge: Large Language Models (LLMs) are undergoing a dynamic transformation . however, there is a potential risk of LLMs creating gossips when prompted with contexts .
Approach: a dataset is used to identify prompts that lead to the creation of gossipy content in the french language.
Outcome: a new dataset identifies prompts that lead to the creation of gossipy content in the french language . the model achieves an accuracy of 89.95% .

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