Raghuveer Thirukovalluru | Duke University

About Me

I am a 5-th year Ph.D. student at Duke University. I am very fortunate to be advised by Prof. Bhuwan Dhingra. Prior to starting my PhD, I was a masters student at UMass Amherst where I worked on amazing projects with IESL. I spent four amazing years of my undergrad at IIT Kanpur.

After my undergrad I spent 2 years at Xerox Research working on some text and graph problems. Over the years, I’ve interned at multiple places including FAIR, MetaAI, XRCI, IESL.

Research Interests

My research in Machine Learning for Natural Language Processing (NLP) focuses on enhancing performance of Large Language Models. I’ve worked on integrating knowledge graphs to improve reasoning in question-answering, optimizing coreference resolution efficiency. Currently, I’m interested in adaptively applying inference-time compute to improve long-form generation, embeddings, and other tasks.

News

[Nov. 2025] Summers 25 internship work with FAIR, ‘TGSIE’ on Arxiv.
[Oct. 2025] ‘ACPO’ accepted as Oral presentaion at IJCNLP-AACL.
[Sept. 2025] ‘B3’ accepted as a Spotlight presentation at NeurIPS 2025
[May. 2025] Presented ‘GenEOL’ at NAACL 2025.
[Dec. 2024] Presented ‘Atomic Self-Consistency’ at EMNLP 2024.
[June. 2024] Presented ‘Summary Transformation for Contrastive Learning’ at NAACL.
[May. 2024] Presented ‘Sequence Reducible Holdout Loss for LM Pretraining’ at COLING.
[Apr. 2024] Presented ‘Atomic Self-Consistency’ at the SouthNLPSymposium (Oral).

Awards

Best Ph.D. Preliminary Research Exam Award, Duke Computer Science (2024–2025).

Talks

[Dec. 2025] ‘B3’ lightning talk at Salesforce booth, NeurIPS 2025.
[Oct. 2025] Invited talk for ‘B3’ at Microsoft Research India
[July. 2025] Presented ‘B3’ in AI Orgs Reading @ Meta AI

Selected Publications

Arxiv

Text-Guided Semantic Image Encoder

Raghuveer Thirukovalluru, Xiaochuang Han, Bhuwan Dhingra, Emily Dinan, Maha Elbayad

Arxiv (Nov 2025)

PDF Code
NeurIPS

Breaking the Batch Barrier (B3) of Contrastive Learning via Smart Batch Mining

Raghuveer Thirukovalluru, Rui Meng, Ye Liu, Karthikeyan K, Mingyi Su, Ping Nie, Semih Yavuz, Yingbo Zhou, Wenhu Chen, Bhuwan Dhingra

NeurIPS 2025

PDF Code
Spotlight at NeurIPS
SOTA on MMEB Leaderboard (May 2025)
IJCNLP

Atomic Consistency Preference Optimization for Long-Form Question Answering

Jingfeng Chen*, Raghuveer Thirukovalluru*, Junlin Wang, Kaiwei Luo, Bhuwan Dhingra (*Equal Author)

IJCNLP-AACL 2025

PDF Code
Oral at IJCNLP-AACL 2025
NAACL

GenEOL: Harnessing the Generative Power of LLMs for Training-Free Sentence Embeddings

Raghuveer Thirukovalluru, Bhuwan Dhingra

In Findings of the North American Association for Computational Linguistics (NAACL 2025)

PDF Code
EMNLP

Atomic Self-Consistency for Better Long Form Generations

Raghuveer Thirukovalluru, Yukun Huang, Bhuwan Dhingra

In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024)

PDF Code
Oral at SouthNLPSymposium 2024
NAACL

SumCSE: Summary as a transformation for Contrastive Learning

Raghuveer Thirukovalluru, Xiaolan Wang, Jun Chen, Shuyang Li, Jie Lei, Rong Jin, Bhuwan Dhingra

In Findings of the North American Association for Computational Linguistics (NAACL 2024)

PDF Code
COLING

Sequence Reducible Holdout Loss for Language Model Pretraining

Raghuveer Thirukovalluru, Nicholas Monath, Bhuwan Dhingra, Sam Wiseman

In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)

PDF Code
ACL

Scaling within document coreference to long texts.

Raghuveer Thirukovalluru, Nicholas Monath, Kumar Shridhar, Manzil Zaheer, Mrinmaya Sachan, and Andrew McCallum.

Findings of the Association for Computational Linguistics, ACL-IJCNLP 2021 (2021)

PDF Code
RepL4NLP

Knowledge Informed Semantic Parsing for Conversational Question Answering

Raghuveer Thirukovalluru*, Mukund Sridhar*, Dung Thai*, Shruti Chanumolu, Nicholas Monath, Sankaranarayanan Ananthakrishnan, Andrew McCallum (*Equal Author)

Workshop on Representation Learning for NLP (RepL4NLP-2021).

PDF Code
RepL4NLP

Simultaneously self-attending to text and entities for knowledge-informed text representations

Dung Thai*, Raghuveer Thirukovalluru*, Trapit Bansal*, Andrew McCallum (*Equal Author)

Workshop on Representation Learning for NLP (RepL4NLP-2021).

PDF Code