Research Mission🧑🏽🔬: Productivity AI
Most current AI agents are trained to improve revenues which don’t usually align with what the users truly value. This makes the decisions made by AI baised and suboptimal. Instead, AI should help you learn, grow, achieve and get things done. Its training objective should be optimizing tasks and rewards that you prefer. I dream of creating a productivity AI assistant that manages and executes your tasks by conversing with you (like J.A.R.V.I.S.). I believe communication is one of the most powerful tools invented by mankind. Giving computers the power of understanding human conversations will automate many repetitive jobs and accelerate innovation.
To pursue this dream, I study and work on open-domain dialog, large langauge models and reinforcment learning. Existing research in robotics 1, embodied AI 2 and large language models 3 have demonstrated that fundamental blocks to build such a machine already exists!
Experience
I am a Research Scientist at Databricks Mosaic Reserach. Previously, I graduated with Computer Science Ph.D. from Georgia Tech, co-advised by Prof. Mark Rield and Prof. Alan Ritter. My primary research focus was on undestanding the mechanisms to enable fine grained controllability of language models. I have published multiple papers in ML and NLP conferences on improving the language models response quality, fluency, and safety (see the list of publications below). Before that, I received a Master’s thesis in Computer Science from The Ohio State University under the guidance of Prof. Alan Ritter. Before that, I was a Research Fellow at Microsoft Research (MSR), India, for one year. I worked with Dr. Monojit Choudhury on word-level language identification of code-switched text. I completed my Bachelor’s Degree in Computer Science and Engineering from the Indian Institute of Technology, Kharagpur.
Contact 📧
The best way to contact me is through email abaheti95@gatech.edu. You can find links to my profiles in other social media and github at the bottom of the page.
Publications 📃
Link to Google Scholar 🎓
Link to Semantic Scholar 🎓
-
🍱 (LoL-RL) Leftover Lunch: Advantage-based Offline Reinforcement Learning for Language Models
Authors: Ashutosh Baheti, Ximing Lu, Faeze Brahman, Ronan Le Bras, Maarten Sap, Mark Riedl
ICLR 2024
PDF Code Poster -
Stanceosaurus: Classifying Stance Towards Multicultural Misinformation
Authors: Jonathan Zheng, Ashutosh Baheti, Tarek Naous, Wei Xu, Alan Ritter
EMNLP 2022
PDF Code -
Just Say No: Analyzing the Stance of Neural Dialogue Generation in Offensive Contexts
Authors: Ashutosh Baheti, Maarten Sap, Alan Ritter, Mark Riedl
EMNLP 2021
PDF Code Slides Poster Talk -
🦠 Extracting a Knowledge Base of COVID-19 Events from Social Media
Authors: Shi Zong, Ashutosh Baheti, Wei Xu, Alan Ritter
COLING 2022
PDF Code -
Fluent Response Generation for Conversational Question Answering
Authors: Ashutosh Baheti, Alan Ritter, Kevin Small
ACL 2020
PDF Code Slides Talk -
Generating More Interesting Responses in Neural Conversation Models with Distributional Constraints
Authors: Ashutosh Baheti, Alan Ritter, Jewei Li, Bill Dolan
EMNLP 2018
PDF Code Slides Poster -
Curriculum design for code-switching: Experiments with language identification and language modeling with deep neural networks
Authors: Ashutosh Baheti, Sunayana Sitaram, Monojit Choudhury, Kalika Bali
ICON 2017
PDF -
Non-linear barrier coverage using mobile wireless sensors
Authors: Ashutosh Baheti, Arobinda Gupta
The 22nd IEEE Symposium on Computers and Communications, 2017
PDF Code
Other Interests/Ideas💡/Activities🚣🏽
Somewhat Explored:
- Cooking 🍳
- Traveling ✈️
- Yoga and meditation 🧘🏽
- Hiking 🥾
- Skating 🛼
- Gaming 🎮
- Mobile game development 👨🏽💻
- Having long conversations about life 🧬, universe 🌌, science 🧪, politics 🗳️ and philosophy
Want to Explore (in Future):
- VR app development 👨🏽💻
- Video Blogging ⏯️
🆓 Tools🛠️ and Extensions🧰 for Productivity
Recommended websites and browser extensions to boost your productivity:
- Notion.so: A database-style note-taking website/app. Helps you organize your thoughts/progress into digital notebook.
- Xmind: A digital notebook to make mind maps and diagrams.
- Grammarly: Free grammar checking app/website/extension.
- Quillbot: Free paraphrasing tool.
- Kami: PDF highlighting browser extension. Saves all highlights and annotations on cloud.
- Dark Reader: Darkify any website using this browser extension to reduce strain on your poor eyes. Remember to often blink and hydrate💧!
- Mailtrack: Free Gmail extension to track sent emails.