Sachin Grover
Hello, I am Sachin Grover, thank you for stopping by!
I am a research scientist at PARC, part of SRI International. My research interests are in LLM based agents, and combining planning and learning techniques/systems. At PARC, I was part of several DARPA projects – SAIL-ON (open-world AI agents), KMASS (knowledge management) and EMHAT (human-ai multi-agent collaboration teams). Previously, I have also worked as an applied scientist intern at Amazon Alexa research on language model quantization and compression techniques.
My Ph.D. was oriented towards designing Human-Aware AI techniques using Automated Task Planning for collaboration in Human-Robot teams and was advised by Prof. Subbarao Kambhampati.
I did my Masters at Arizona State University, advised by Prof. Kurt VanLehn, on modeling student’s knowledge while working on Intelligent Tutoring System.
Currently, I am looking for full-time opportunities in research and engineering roles.
sachin . grover @ asu . edu (remove spaces)
Research Interest: LLM Agents, Learning+Planning systems, Neurosymbolic techniques.
Research
My research focuses on designing human-aware techniques and systems utilizing recent improvements in LLMs, planning techniques, and other learning methods. I am also interested in solving real-world complex problems and creating interpretable systems that provide explanations with guarantees utilizing machine learning and AI-planning techniques. My work in the past has been published in refereed journals (AIJ, HCI, IEEE, Taylor & Francis, etc.) and several peer-reviewed conferences (International Conference of Automated Planning and Scheduling ICAPS, Human-Robot Interaction HRI, AI & Education AIED, Naturalistic Decision Making NDM, etc.) and workshops (Knowledge Engineering for Planning and Scheduling ICAPS KEPS, Explainability and AI ICAPS XAIP, ML for Public Health NEURIPS MLPH), etc. A list of my publications can be checked below.
publications
2024
- AIJA domain-independent agent architecture for adaptive operation in evolving open worldsArtificial Intelligence by Elsevier , 2024
- AAAI SymposiumSelf-monitoring Adaptive AI Agents Operating in Open WorldsAAAI Spring Symposium on User-Aligned Assessment of Adaptive AI Systems., 2024
- ICAPS KEPSNyx: Domain Independent PDDL+ planner for Classic Control ProblemsICAPS workshop on Knowledge Engineering for Planning and Scheduling. Also submitted to KR conference, 2024
- ICAPS DemoSelf-adaptive Mission Planning in High Fidelity Open World SimulationICAPS Demonstration, 2024
- ICAPS DemoA Demonstration of Natural Language Understanding for Embodied Agent using LLMsICAPS Demonstration, 2024
2023
- ICAPSHeuristic search for physics-based problems: angry birds in PDDL+Proceedings of the International Conference on Automated Planning and Scheduling, 2023
2022
- Doctoral ThesisHuman Aware AI Methods for Active TeamingArizona State University , 2022
2021
- NEURIPS MLPHCOVID-19 India Dataset: Parsing COVID-19 Data in Daily Health Bulletins from States in IndiaNeurips MLPH, arXiv preprint arXiv:2110.02311, 2021
- PatentCompression of machine learned modelsApr 2021US Patent 10,970,470
2020
- HCI special issueRADAR: automated task planning for proactive decision supportHuman–Computer Interaction by Taylor & Francis , Apr 2020
- ICAPS XAIPModel elicitation through direct questioningICAPS XAIP, arXiv preprint arXiv:2011.12262, Apr 2020
- PatentCompression of machine learned modelsFeb 2020US Patent 10,558,738
2019
- HRIPlan Explanations as Model Reconciliation – An Empirical StudyIn 2019 14th ACM/IEEE International Conference on Human-Robot Interaction (HRI) , Feb 2019
- NDMiPass: A case study of the effectiveness of automated planning for decision supportNaturalistic Decision Making, Feb 2019
2018
- ICAPS SPARKWhat can automated planning do for intelligent tutoring systemsICAPS SPARK, Feb 2018
- IJAIEDHow should knowledge composed of schemas be represented in order to optimize student model accuracy?In International Conference on Artificial Intelligence in Education , Feb 2018
2017
- IEEELearning How to Construct Models of Dynamic Systems: An Initial Evaluation of the Dragoon Intelligent Tutoring SystemIEEE Transactions on Learning Technologies, Feb 2017
- The design and development of the dragoon intelligent tutoring system for model construction: lessons learnedInteractive Learning Environments by Routledge , Feb 2017
2016
- IJAIEDLearning science by constructing models: can dragoon increase learning without increasing the time required?International Journal of Artificial Intelligence in Education by Springer , Feb 2016
2015
- Masters ThesisOnline Embedded Assessment for Dragoon, Intelligent Tutoring SystemArizona State University , Feb 2015
2009
- IEEEText Extraction from Document Images Using Edge InformationIn Annual IEEE India Conference , Feb 2009