Abdul Waheed

I am a PhD student at School of Computer Science, Carnegie Mellon University in Pittsburgh, PA.

My research focuses on building a principled understanding of reasoning in foundation models: how reasoning emerges, how to induce it efficiently [NeurIPS'25 Workshops], and how the structure and the format of data shapes it [ICLR'26]. I am interested in optimal reasoning depth, the structural conditions that enable systematic generalization, and reliable evaluation of model capabilities [ICLR'26]. In the past, I have extensively worked on evaluation [Interspeech'23, EMNLP'23, ACL'25, ACL'25] and efficiency [EACL'24, ACL'24, NAACL'25].

Outside of research, I enjoy playing badminton, reading, hiking, and traveling. I like trees, rain, and preferably both together.

Email / Scholar / Twitter/X / Github / Blog

Want to talk? I have learned a lot from my advisors, mentors, collaborators, and friends who have supported me along the way. I'm happy to share what I've learned and meet others with similar interests. If you'd like to discuss research, explore a collaboration, or talk about applying to PhD programs, feel free to reach out.

News

Feb 2026VideoJudge and Code-Induced Reasoning accepted at ICLR 2026.
May 2025Two papers accepted at ACL 2025.
Feb 2025New preprints: Approximating ASR Metrics and Speech Hallucinations.
Jan 2025uDistil-Whisper accepted at NAACL 2025 (main).
Oct 2024New preprints: uDistil-Whisper, Speech Foundation Models, and Synthetic Data Diversity.
Aug 2024Joined LTI at Carnegie Mellon University as a graduate student.
Jun 2024One main conference paper and one workshop paper accepted at ACL 2024.

Selected Papers

* denotes equal contribution / first co-authors

Please visit scholar page for more details.

	Encoder-Decoder Manifold Alignment for Idempotent Generation Dareen Alharthi, Ibrahim Aldarmaki, Abdul Waheed, Semyon Semenov, Bhiksha Raj *DynaFront Workshop, NeurIPS, 2026* Currently under review Paper
	VideoJudge: Bootstrapping Enables Scalable Supervision of MLLM-as-a-Judge for Video Understanding Abdul Waheed , Zhen Wu, Dareen Alharthi, Seungone Kim, Bhiksha Raj *ICLR, 2026* Paper / Code
	On Code-Induced Reasoning in LLMs Abdul Waheed* , Zhen Wu, Carolyn Rosé, Daphne Ippolito ICLR, 2026* Paper / Code
	Less is More Tokens: Efficient Math Reasoning via Difficulty-Aware Chain-of-Thought Distillation Abdul Waheed, Chancharik Mitra, Laurie Z. Wang, Deva Ramanan, Bhiksha Raj Efficient Reasoning and Multimodal Algorithmic Reasoning workshops at NeurIPS*, 2025 Paper / Code
	Lost in Transcription, Found in Distribution Shift: Demystifying Hallucination in Speech Foundation Models Hanin Atwany, Abdul Waheed , Rita Singh, Monojit Choudhury, Bhiksha Raj *ACL, 2025* Paper / Code
	On the Robust Approximation of ASR Metrics Abdul Waheed , Hanin Atwany, Rita Singh, Bhiksha Raj *ACL, 2025* Paper / Code
	On the Diversity of Synthetic Data and its Impact on Training Large Language Models Hao Chen, Abdul Waheed , Xiang Li, Yidong Wang, Jindong Wang, Bhiksha Raj, Marah I. Abdin *Preprint, 2024* Paper
	What Do Speech Foundation Models Not Learn About Speech? Abdul Waheed , Hanin Atwany, Bhiksha Raj, Rita Singh *Preprint, 2024* Paper
	uDistil-Whisper: Label-Free Data Filtering for Knowledge Distillation in Low-Data Regimes Abdul Waheed , Karima Kadaoui, Muhammad Abdul-Mageed *NAACL Main, 2025* Paper / Code
	To Distill or Not to Distill? On the Robustness of Robust Knowledge Distillation Abdul Waheed , Karima Kadaoui, Muhammad Abdul-Mageed *ACL Main, 2024* Paper \| Models
	Towards Zero-Shot Text-To-Speech for Arabic Dialects Khai Duy Doan, Abdul Waheed , Muhammad Abdul-Mageed *ArabicNLP conference co-located with ACL , 2024* Paper
	LaMini-LM: A Diverse Herd of Distilled Models from Large-Scale Instructions Minghao Wu, Abdul Waheed , Chiyu Zhang, Muhammad Abdul-Mageed, Alham Fikri Aji *EACL Main, 2024* Paper \| Models \| Data
	GPTAraEval: A Comprehensive Evaluation of ChatGPT on Arabic NLP Md Tawkat Islam Khondaker, Abdul Waheed , El Moatez Billah Nagoudi, Muhammad Abdul-Mageed *EMNLP Main, 2023* Paper
	TARJAMAT: Evaluation of Bard and ChatGPT on Machine Translation of Ten Arabic Varieties Karima Kadaoui, Samar M. Magdy, Abdul Waheed* , Md Tawkat Islam Khondaker, Ahmed Oumar El-Shangiti, El Moatez Billah Nagoudi, Muhammad Abdul-Mageed *ArabicNLP conference co-located with EMNLP, 2023* Paper
	N-Shot Benchmarking of Whisper on Diverse Arabic Speech Recognition Bashar Talafha, Abdul Waheed , Muhammad Abdul-Mageed *Interspeech, 2023* Paper \| Code
	Speaker and Time-aware Joint Contextual Learning for Dialogue-act Classification in Counselling Conversations Ganeshan Malhotra, Abdul Waheed , Aseem Srivastava, Md Shad Akhtar, Tanmoy Chakraborty *ACM International Conference on Web Search and Data Mining (WSDM), 2022* Paper
	Analyzing the Domain Robustness of Pretrained Language Models, Layer by Layer Abhinav Ramesh Kashyap, Laiba Mehnaz, Bhavitvya Malik, Abdul Waheed , Devamanyu Hazarika, Min-Yen Kan, Rajiv Shah *Domain Adaptation for NLP workshop co-located with EACL, 2021* Paper

Fork from Jon Barron's website.