Pulkit Agrawal
Associate Professor
Studies robot learning, including reinforcement learning, dexterous manipulation, and self-supervised learning of sensorimotor skills.
Pulkit Agrawal's lab at MIT works on manipulation & robot learning, reinforcement learning and generative & foundation models. 60 papers since 2023, 49 at top venues, mostly ICML and ICRA. 12 PhD students. 48 alumni, most often to a Stealth AI Startup. and a Stealth Startup. Applications close .
At a glance
- papers, 2023–2026
- 16/ year
- at T1 venues
- 82%
- PhD students
- 12
- MS students
- 0
- alumni listed
- 48
Papers by yearfrom the lab homepage, 2023–2026
- 0 papers2026 · T1 0 · collecting
- 19 papers2025 · T1 12
- 21 papers2024 · T1 19
- 20 papers2023 · T1 18
T1 venuesOtherVenues: ICML, ICRA, NeurIPS, ICLR, CoRL
Research focuspaper titles that use each topic's words
- Manipulation & robot learning34 papers
- Reinforcement learning29 papers
- Generative & foundation models6 papers
- skills1 paper
- dexterous manipulation1 paper
Paper list
From the lab homepage.
202519 papers
- npj Digital MedicineA distributional reinforcement learning model for optimal glucose control after cardiac surgery (opens in a new tab)
- Nature (T1)A multimodal robotic platform for multi-element electrocatalyst discovery (opens in a new tab)
- RSS (T1)Bridging the Sim-to-Real Gap for Athletic Loco-Manipulation (opens in a new tab)
- ICRA (T1)DexHub and DART: Towards Internet Scale Robot Data Collection (opens in a new tab)
- ICLR (T1)Diffusion Policy Policy Optimization (opens in a new tab)
- ICLR (T1)Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning (opens in a new tab)
- ICML (T1)Emergence and Effectiveness of Task Vectors in In-Context Learning: An Encoder Decoder Perspective (opens in a new tab)
- ICRA (T1)From Imitation to Refinement – Residual RL for Precise Visual Assembly (opens in a new tab)
- ICML (T1)General Intelligence Requires Reward-Based Pretraining (opens in a new tab)
- RoboSoftHovering Flight of Soft-Actuated Insect-Scale Micro Aerial Vehicles using Deep Reinforcement Learning (opens in a new tab)
- Journal of the American Society of NephrologyKidneyPilot: Reinforcement Learning Model for Prevention of Persistent AKI After Cardiac Surgery
- npj Computational MaterialsKnown Unknowns: Out-of-Distribution Property Prediction in Materials and Molecules (opens in a new tab)
202421 papers
- CoRL (T1)Action Space Design in Reinforcement Learning for Robot Motor Skills (opens in a new tab)
- HRI (T1)Aligning human and robot representations (opens in a new tab)
- ICLR (T1)Curiosity-driven Red-teaming for Large Language Models (opens in a new tab)
- ICRA (T1)Everyday finger: a robotic finger that meets the needs of everyday interactive manipulation (opens in a new tab)
- IROS (T1)EyeSight Hand: Design of a Fully-Actuated Dexterous Robot Hand with Integrated Vision-Based Tactile Sensors and Compliant Actuation (opens in a new tab)
- NeurIPS (T1)Few-Shot Task Learning through Inverse Generative Modeling (opens in a new tab)
- NeurIPS (T1)Going Beyond Heuristics by Imposing Policy Improvement as a Constraint (opens in a new tab)
- TMLR (T1)Grid Cell-Inspired Fragmentation and Recall for Efficient Map Building (opens in a new tab)
- IROS (T1)JUICER: Data-Efficient Imitation Learning for Robotic Assembly (opens in a new tab)
- ICRA (T1)Learning Force Control for Legged Manipulation (opens in a new tab)
- NeurIPS (T1)Learning Multimodal Behaviors from Scratch with Diffusion Policy Gradient (opens in a new tab)
- ICRA (T1)Lifelong Robot Learning with Human Assisted Language Planners (opens in a new tab)
202320 papers
- CoRL (T1)Autonomous Robotic Reinforcement Learning with Asynchronous Human Feedback (opens in a new tab)
- NeurIPS (T1)Beyond Uniform Sampling: Offline Reinforcement Learning with Imbalanced Datasets (opens in a new tab)
- NeurIPS (T1)Breadcrumbs to the Goal: Goal-Conditioned Exploration from Human-in-the-Loop Feedback (opens in a new tab)
- NeurIPS (T1)Compositional Foundation Models for Hierarchical Planning (opens in a new tab)
- ICML (T1)Diagnosis, Feedback, Adaptation: A Human-in-the-Loop Framework for Test-Time Policy Adaptation (opens in a new tab)
- International Conference on Robotics and Automation (ICRA)DribbleBot: Dynamic Legged Manipulation in the Wild (opens in a new tab)
- ICLR (T1)Harnessing Mixed Offline Reinforcement Learning Datasets via Trajectory Weighting (opens in a new tab)
- ICLR (T1)Is Conditional Generative Modeling all you need for Decision Making? (opens in a new tab)
- ICLR (T1)Learning to Extrapolate: A Transductive Approach (opens in a new tab)
- Conference on Robot Learning (CoRL)Learning to See Physical Properties with Active Sensing Motor Policies (opens in a new tab)
- ICML (T1)Parallel Q-Learning: Scaling Off-policy Reinforcement Learning under Massively Parallel Simulation (opens in a new tab)
- NeurIPS (T1)Self-Supervised Reinforcement Learning that Transfers using Random Features (opens in a new tab)
Where alumni went48 listed, 44% to industry
Industry 8Academia 10
Most often to
- a Stealth AI Startup.
- a Stealth Startup
Also to
- Stanford University
- Carnegie Mellon University
- ETH Zurich
Apply to MITFall 2027
MIT EECS Computer Science PhD
- Round
- Single round
- Opens
- —
- GRE
- See program page
- English
- TOEFL/IELTS
- Fee
- USD 90
Confirmed on the program page.
Similar labsnearest by research summary
- USCMU
Changliu Liu
- Human-robot interaction & assistive robots
- SGNUS
Guillaume Sartoretti
- Planning, control & multi-robot systems
- USGeorgia Tech
Gil Weinberg
1 T1 - SGNTU
Lihua Xie
- Planning, control & multi-robot systems
- USCMU
Jeff Schneider
- Autonomous vehicles & drones
- USUMD
Calin Belta
- Planning, control & multi-robot systems
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
MITcloses