Brihi Joshi

I am a third-year CS PhD Candidate at the University of Southern California, advised by professors Xiang Ren and Swabha Swayamdipta as a part of the NLP Group, INK Lab and Dill Lab. My research interests lie in LLM explainability and human-centered operationalization of LLMs – how can humans effectively use LLMs and in turn, how can LLMs benefit from human intervention. I often use rationales or model explanations as the medium of communication between LLMs and humans. My research is supported by the USC Annenberg, 2023 Amazon ML and 2024 Apple Scholars in AI/ML Fellowships.

Prior to this, I was an Analyst at Goldman Sachs where I worked with the Regulatory Engineering team. Even before that, I graduated with a Bachelors in Computer Science from IIIT Delhi in December, 2020, where I completed my thesis as a part (and founding member) of the Laboratory for Computational Social Systems (LCS2).

Apart from pursuing my academic interests, I am extremely passionate about serving the community. I used to co-lead the Delhi chapter of Women Who Code Delhi, which is an organisation that works towards diversity and inclusion of women in tech, for more than 5 years. Currently, I am one of the organizers of the NLP with Friends, which is an online NLP seminar for and by students. I am a huge advocate of diversity in CS Research, and always looking for opportunities to support them! Head over to the outreach & mentorship tab for more details! I am also a musician by passion and absolutely love singing and composing music. I’ve tried my hands at 3 instruments (and I plan to master some, explore more).


May 2024 Started as an ML Research Intern at Apple in Seattle! Super excited to work on interesting problems in Humans x LLMs 🙌🏽
Mar 2024 Awarded the 2024 Apple Scholars in AI/ML PhD fellowship! 🥺
Jan 2024 MaRio accepted at ICLR 2024! See you in Vienna!! 🙌🏽
Nov 2023 Excited to share our new work led by Sahana, on improving the quality of rationales generated by small LMs!
Oct 2023 Passed my PhD Quals! Officialy a PhD Candidate now :D
Oct 2023 Gave a talk at the USC-Amazon Centre on Trusted AI about Human-Centred Operationalization of Model Explanations!
Jul 2023 Honoured to be awarded the Amazon ML PhD Fellowship for 2023-24! 🥺
May 2023 Started as an Applied Scientist Intern at Amazon Alexa 🎉 Excited to explore the greater Boston area and work with the NLG Team!
May 2023 XMD got accepted to ACL Demo Track 2023! 🎉
May 2023 Our human utility project got accepted to ACL (Main Conference) 2023! See you in Toronto! 🎉🍾
Dec 2022 Excited to share a new preprint on measuring human utility of free-text rationales, that I co-led with Ziyi!
Dec 2022 New preprint, 🔪 KNIFE, on learning from free-text rationales is out! Happy to have collaborated with Aaron and Zhiyuan who led the work!
Oct 2022 ER-Test is now accepted at EMNLP 2022 as a Findings Paper! I will be presenting it at the BlackboxNLP workshop on 12/08!
Oct 2022 Our proposal with Aaron and Xiang, on Utilizing Explanations for Model Refinement got the Alexa: Fairness in AI award!
Aug 2022 New demo paper XMD out!
🕰️ all news...