Carnegie Mellon University
Arnav Goel
Hello! I am a graduate student in the Machine Learning Department at Carnegie Mellon University, where I am a research assistant with Dr. Daniel Fried and Dr. Sean Welleck.
I am currently an Applied Scientist Intern at AWS Agentic AI, working on reliable production-grade multi-agent systems.
Research
My research interests lie in reliable machine learning. Recent work spans domain-specific verification, multi-agent reliability, model post-training, and token-efficient reasoning in large language models.
Prior work covers distillation of large models into efficient deployable ones, attribution of model behaviour to pre-training corpora, alignment and safety for enterprise language models, temporal and long-horizon user modelling, and multilingual speech and vision-language systems.
About
I have had the privilege of working with Dr. Aditi Raghunathan at CMU, Dr. Amit Sharma at Microsoft Research India, Dr. Xiang Ren at USC’s INK Lab, Dr. Jackie Chi Kit Cheung at Mila and McGill, and Dr. Anubha Gupta and Dr. Rajiv Ratn Shah during my undergraduate degree at IIIT Delhi. The internships at USC and Mila were funded by the IUSSTF-Viterbi Research Program 2024, for which I was selected among 15 students nationwide, and the MITACS Globalink Research Scholarship.
Outside research I am a big-time history buff and love watching documentaries. I have recently developed a strong penchant for cooking, and swimming has always been my way of unwinding. Please feel free to reach out — I am always glad to talk about language models and research more broadly, and especially happy to advise students finding their way into the field.
Selected publications all publications →
HORIZON: A Benchmark for In-the-wild User Behaviour Modeling
Findings of ACL 2026 · NeurIPS 2025 Workshop
Exploring Multilingual Unseen Speaker Emotion Recognition: Leveraging Co-Attention Cues in Multitask Learning
INTERSPEECH 2024
News
- May 2026Started as an Applied Scientist Intern with the AWS Agentic AI team in Seattle for the summer.
- Apr 2026Two papers — Mind the Gap and MORPHOGEN — accepted to ACL 2026.
- Apr 2026HORIZON, the user-modeling benchmark I built at MSR, accepted to Findings of ACL 2026.
- Feb 2026Preprint out on On Defining Erasure Harms for NLP, joint work with Mila, McGill and the FATE team at Microsoft Research Montréal.
- Oct 2025HORIZON accepted at a NeurIPS 2025 workshop.
- Aug 2025Started my M.S. in Machine Learning at Carnegie Mellon. Excited to be in Pittsburgh!
- Jan 2025Paper on attributing culture-conditioned generations to pretraining corpora accepted at ICLR 2025 — my work with the INK Lab at USC.
- Jan 2025Selected for the Google DeepMind Research Symposium 2025. Looking forward to connect with people in Bangalore!
- Jan 2025Started as a Research Engineering Intern at Microsoft Research India with the Extreme Classification Group.
- Oct 2024My team DBkaScam was placed 6th amongst 20,000+ teams in the Amazon ML Challenge 2024! Check out our solution at this link.
- Jul 2024Started my MITACS Globalink research internship at Mila-Quebec AI Institute and McGill University.
- Jun 2024Full paper on multilingual unseen speaker emotion recognition accepted at INTERSPEECH 2024, co-led as part of my undergraduate thesis at IIIT-Delhi.
- May 2024Thrilled to share that I have started my internship at the INK Lab in the USC NLP Group as one of the 15 IUSSTF-Viterbi Research Scholars selected from all over India. Looking forward to an incredible summer in Los Angeles :)
- Feb 20242 papers on speech processing and multilingual prosody transfer accepted at the Tiny Track at ICLR 2024!
- Nov 2023LLMGuard accepted at the Demonstrations Track at AAAI 2024, from my internship at IBM Research India.
- Aug 20232 papers on improving LLMs in scientific question answering accepted at BDA 2023, from the Summer Undergraduate Research Fellowship at IIIT-Delhi.