Yash Kumar Lal

Yash Kumar Lal


PhD student in Computer Science at Stony Brook; affiliated with LUNR lab

Advisor: Niranjan Balasubramanian.

Email: ylal <at> cs.stonybrook <dot> edu

Hello!

Hi! I'm a fifth-year PhD student at Stony Brook. I work on implicit reasoning problems on diverse types of texts. Particularly, I focus on evaluating and improving the abilities of NLP models to answer why questions that elicit different aspects of reasoning on stories and plans. I collaborate extensively with Nate Chambers and Ray Mooney on this line of work. I am also interested in understanding how well large language models (LLMs) understand the social and personal aspects of reasoning expressed in language, working with H. Andrew Schwartz.

Previously, I have dabbled with problems in analyzing energy efficiency of pretrained models, machine translation, clickbait detection and word sense disambiguation.

I graduated with a Master's degree in Computer Science from Johns Hopkins University in May 2020. I was primarily advised by Philipp Koehn and worked on a variety of problems across natural language processing.

Over the years, I have been involved in several notable side projects. I worked on a chat platform - Ping - that allowed users to communicate with each other regardless of language. I was involved in efforts to build a service - hello friend - that allowed people without internet access to avail various crucial facilities. In a past life, I was an iOS developer for several small-scale applications.


Updates

October 2024

Presented CaT-Bench at the NYC ML Symposium organized by the New York Academy of Sciences.

September 2024

Long paper accepted to EMNLP, 2024. I am also co-organizing WNU 2024. I'll be attending in-person. See you in sunny Miami!

May 2024

Paper from my AI2 Internship accepted to ACL (Findings), 2024. I will be presenting it at the NLRSE workshop. See you in Bangkok!

April 2024

Paper from my Google Research Internship accepted to TrustNLP workshop collocated with NAACL, 2024. I'll be presenting in Mexico City!

January 2024

Short paper accepted to EACL, 2024.

August 2023

I am spending the Fall semester at Google Research with the Responsible, Application-Driven Machine Learning (RADML) team in the Responsible AI and Human Centered Technology (RAI-HCT) org working with Ananth Balashankar, Ahmad Beirami and Preethi Lahoti.

May 2023

Short paper accepted to ACL, 2023. Another accepted to a workshop (WASSA). My co-authors will be presenting in Toronto!

February 2023

I am spending the summer at Allen Institute for Artificial Intelligence in the Aristo team working with Niket Tandon.

October 2022

Long paper accepted to EMNLP, 2022. I'll be attending in-person. See you in Abu Dhabi!

May - August 2022

I am spending the summer working on question decomposition for multi-hop question answering at Salesforce Research with Semih Yavuz and Yingbo Zhou.

May 2022

Presenting work on using commonsense knowledge to answer why questions in stories at WNU 2022. See you in Seattle!

May 2021

Two papers accepted at ACL, 2021 --- one in main conference, one in Findings


Publications

  • CaT-BENCH: Benchmarking Language Model Understanding of Causal and Temporal Dependencies in Plans
    In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024 [Online] [Paper] [BibTex] [Poster] [Slides] [Dataset]
  • Tailoring with Targeted Precision: Edit-Based Agents for Open-Domain Procedure Customization
    In Proceedings of the Findings of the Association for Computational Linguistics, ACL 2024 [Online] [Paper] [BibTex]
  • Automated Adversarial Discovery for Safety Classifiers
    Proceedings of the 4rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2024), at NAACL [Online] [Paper] [BibTex] [Poster] [Slides] Runner-Up, Best Long Paper
  • SOCIALITE-LLAMA: An Instruction-Tuned Model for Social Scientific Tasks
    In Proceedings of the Association for Computational Linguistics, EACL 2024 [Online] [Paper] [BibTex] [Code]
  • SAGE-viz: SchemA GEneration and Visualization
    In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, EMNLP 2023 [Online] [Paper] [BibTex]
  • Systematic Evaluation of GPT-3 for Zero-Shot Personality Estimation
    In Proceedings of the 13th Workshop on Computational Approaches to Subjectivity, Sentiment, & Social Media Analysis (WASSA 2023), at ACL [Online] [Paper] [BibTex]
  • Evaluating Paraphrastic Robustness in Textual Entailment Models
    In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics, ACL 2023 [Online] [Paper] [BibTex] [Code]
  • Using Commonsense Knowledge to Answer Why-Questions
    In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022 [Online] [Paper] [BibTex] [Slides] [Code]
  • IrEne-viz: Visualizing Energy Consumption of Transformer Models
    In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, EMNLP 2021 [Paper] [BibTex] [Poster] [Demo]
  • IrEne: Interpretable Energy Prediction for Transformers
    In Proceedings of the Association for Computational Linguistics, ACL 2021 [Online] [Paper] [BibTex]
  • [Code]
  • TellMeWhy: A Dataset for Answering Why-Questions in Narratives
    In Proceedings of the Findings of the Association for Computational Linguistics, ACL 2021 [Online] [Paper] [BibTex] [Poster] [Slides] [Dataset] [Code]
  • Temporal Reasoning in Natural Language Inference
    In Proceedings of the Findings of the Association for Computational Linguistics, EMNLP 2020 [Online] [Paper] [BibTex]
  • Sentence-Level Adaptation for Low-Resource Neural Machine Translation
    In Proceedings of the AMTA 2019 Workshop on Technologies for MT of Low Resource Languages (LoResMT) 2019 [Online] [Paper] [BibTex]
  • De-Mixing Sentiment from Code-Mixed Text
    In Proceedings of the 57th Annual Meeting of Association for Computational Linguistics - Student Research Workshop (ACL-SRW) 2019 [Online] [Paper] [BibTex]
  • Johns Hopkins University Submission for WMT News Translation Task
    In Proceedings of the Fourth Conference on Machine Translation (WMT) 2019 [Online] [Paper] [BibTex]
  • Check It Out : Politics and Neural Networks
    In Proceedings of CLEF 2018 Fact-Checking Shared Task
  • Identifying Clickbait: A Multi-Strategy Approach Using Neural Networks
    In Proceedings of The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval, (SIGIR) 2018 [Online] [Paper] [BibTex]
  • SWDE: A Sub-Word And Document Embedding Based Engine for Clickbait Detection
    In Proceedings of SIGIR 2018 Workshop on Computational Surprise in Information Retrieval, (CompS Workshop) [Online] [Paper] [BibTex]

Service