←
Learning and Intelligent Systems (LIS) @ MIT142 posts
Learning and Intelligent Systems (LIS) @ MIT avatar

Learning and Intelligent Systems (LIS) @ MIT

@MIT_LISLab

PIs: Leslie Pack Kaelbling, Tomás Lozano-Pérez AI/ML/Robotics Research @MIT_CSAIL Website: lis.csail.mit.edu

◷ Joined November 2020▧ 14 Photos and videos
Weirui Ye avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
Weirui Ye@Weirui_Ye·Nov 8

Planners operate on task timescales. Feasibility changes on control timescales. Bridging that gap takes a high-frequency mid-level layer. Introducing EPO: Energy-Based Policy Orchestration #CoRL2026 EPO gives each low-level policy an energy: how feasible a goal is, and how well the policy will reach it. Energies are differentiable and comparable across policies, so EPO optimizes the policy–goal pair at control rate. Unlike a learned MoE router, EPO is compositional: adding a skill means training one new energy, nothing else. It's also more precise and switches policies 7× less often, pushing the precision–robustness frontier. 📄 Paper: arxiv.org/abs/2611.02546 🌐 Project: yewr.github.io/epo 💻 Code: github.com/YeWR/EPO @wangyenjen @elvis_hsieh77 @frankzydou @ShiboZhaoSLAM @pabbeel @du_yilun @MIT_LISLab

0:51
Learning and Intelligent Systems (LIS) @ MIT avatar
Learning and Intelligent Systems (LIS) @ MIT@MIT_LISLab·Nov 7

Can’t find something? Come ask our robots for help!

Imagine asking your robot to find your car keys🔑. They could be on your desk, the key holder, or in a drawer. The robot can't see most of the room - it has to reason about likely hiding spots, go look, and open things. Can robots do this today? Our new work, SSF, does exactly that! 🤖🧵

0:25
Yifan Kang @ CoRL 2026 avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
Yifan Kang @ CoRL 2026@kyiiiifan·Nov 5

Excited to share my first project at @MIT_LISLab : 𝐊𝐞𝐲𝐅𝐥𝐨𝐰! 🤖 Can robots adapt from a few demos without retraining? We let a coding agent improve a retrieval-based policy through simulated play. Presenting at CoRL Learn@Deploy workshop! 🧵(1/7) yifank.github.io/keyflow/

0:35
Learning and Intelligent Systems (LIS) @ MIT avatar
Learning and Intelligent Systems (LIS) @ MIT@MIT_LISLab·Jun 3

Exciting to see the launch of the NVIDIA Isaac GR00T Reference Humanoid Robot. Open, research-focused humanoid platforms can help accelerate progress across the robotics community. Looking forward to the innovations this enables. A great contribution from NVIDIA Robotics!

NVIDIA announces the first open humanoid robot reference design built for robotics research. The NVIDIA Isaac GR00T Reference Humanoid Robot combines the @UnitreeRobotics H2 humanoid robot, @SharpaRobotics Wave five-fingered hands for dexterous manipulation, Jetson Thor onboard compute, and Isaac GR00T open software and models, giving researchers a full-stack platform from data capture to model deployment. Read the #NVIDIAGTC Taipei announcement: nvda.ws/4ef9VOr

0:20
Aditya Agarwal avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
Aditya Agarwal@skymanaditya1·Jun 2

Excited to be at #ICRA2026 in Vienna this week 🇦🇹 -- I'll be presenting SCOUT as a spotlight talk at the Rigorous Robot Perception workshop on June 01 at VIP Lounge C, and SceneComplete as an Oral talk today (June 02 at TuAT3.5).

Post attachmentPost attachmentPost attachmentPost attachment
William Shen avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
William Shen@WillShenSaysHi·May 8

𝗧𝗶𝗣𝗧𝗼𝗣 𝗶𝘀 #𝟭 𝗼𝗻 𝗠𝗼𝗹𝗺𝗼𝗦𝗽𝗮𝗰𝗲𝘀! Outperforming VLAs including MolmoAct2 and π₀.₅, and WAMs like DreamZero It's the only method that uses inference-time search and 𝙯𝙚𝙧𝙤 robot data. We didn't do any benchmark-specific tuning.

Post attachment
Nishanth Kumar (✈️ @ CoRL 2026) avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
Nishanth Kumar (✈️ @ CoRL 2026)@nishanthkumar23·May 8

We ran TiPToP on MolmoSpaces and achieved SOTA! And we wrote an X article thing to discuss details - excited to hear any questions/comments! Check out the leaderboard here: molmospaces.allen.ai/leaderb…. TiPToP is also fully-open and should take <1hr of setup: try it yourself here: tiptop-robot.readthedocs.io/…

Nishanth Kumar (✈️ @ CoRL 2026) @nishanthkumar23 May 8 Article Achieving SOTA on the MolmoSpaces benchmark with Inference-Time Search We recently developed TiPToP: a general-purpose manipulation system that uses foundation models for perception followed by inference-time search for trajectory generation (specifically via Task and

Learning and Intelligent Systems (LIS) @ MIT avatar
Learning and Intelligent Systems (LIS) @ MIT@MIT_LISLab·Apr 24

Congrats to the newest PhD grad from the lab @nishanthkumar23 🤖

PhDone and officially a robot doctor! Thanks to incredible advisors Leslie and Tomás @MIT_LISLab and my secret bonus advisor @tomssilver - I’m so so lucky to have gotten to do research with them. I’m excited to keep working on making useful generalist robots a reality 🤖🚀

Post attachment
Nishanth Kumar (✈️ @ CoRL 2026) avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
Nishanth Kumar (✈️ @ CoRL 2026)@nishanthkumar23·Apr 6

We've released the TiPToP code open source! 🚀 We put a lot of effort into making this easy to run, install and build on. You can get it running on your DROID setup in under an hour. No training data needed. No demonstrations. Just type in what you want your robot to do 🤖💬 You can also adapt it to new embodiments with minimal effort. Give it a try and let us know what you think! 💻 Code: github.com/tiptop-robot/tipt… 📖 Docs: tiptop-robot.readthedocs.io/

State-of-the-art robot policies often need hundreds of hours of data. What if we needed none? Introducing TiPToP: a manipulation system that zero-shots open-world tasks from pixels and language using vision foundation models and GPU-parallelized Task and Motion Planning (TAMP).

3:43
Lucy Cai avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
Lucy Cai@lucyjcai·Mar 18

Imagine you told a robot to "find your car keys" in your apartment and it looked around, opened a drawer, and retrieved them for you. As a step towards that, I adapted TiPToP to run on the RBY1 humanoid in our lab! Here's an example instruction it follows: "Put the green block on the blue plate and the yellow block on the magazine." TiPToP helps plan over the right arm + single torso joint, but it's easy to unlock more joints -- even the base wheels -- for more expressive, real-world tasks. Humans find objects without thinking twice. One day, robots will too! 🤖

0:40

State-of-the-art robot policies often need hundreds of hours of data. What if we needed none? Introducing TiPToP: a manipulation system that zero-shots open-world tasks from pixels and language using vision foundation models and GPU-parallelized Task and Motion Planning (TAMP).

3:43
Robots Digest 🤖 avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
Robots Digest 🤖@robotsdigest·Mar 12

Everyone is scaling VLAs with more robot data. TiPToP shows another path. No robot training, no policy learning. Just RGB + language → 3D scene → GPU TAMP planner → trajectory. Foundation models + planning alone can run real manipulation tasks.

1:39
Learning and Intelligent Systems (LIS) @ MIT avatar
Learning and Intelligent Systems (LIS) @ MIT@MIT_LISLab·Mar 11

Check out a new robot manipulation system from members of the group combining foundation models with TAMP!

State-of-the-art robot policies often need hundreds of hours of data. What if we needed none? Introducing TiPToP: a manipulation system that zero-shots open-world tasks from pixels and language using vision foundation models and GPU-parallelized Task and Motion Planning (TAMP).

3:43
Ken Goldberg avatar
⟳ Learning and Intelligent Systems (LIS) @ MIT retweeted
Ken Goldberg@Ken_Goldberg·24 Nov 2025

Data Flywheel -> Data Avalanche: Thx to Leslie Kaebling and @Pulkitology Agrawal for suggesting "avalanche" as a better metaphor than "flywheel" for combining model-free + model-based methods to bootstrap a specific robot task to bootstrap & amplify on-policy data collection.

Post attachment
Learning and Intelligent Systems (LIS) @ MIT avatar
Learning and Intelligent Systems (LIS) @ MIT@MIT_LISLab·25 Nov 2025

Check out new work from @nishanthkumar23 in the group on learning and planning with symbolic world models!

RoboPapers@RoboPapers24 Nov 2025

Reasoning over long horizons would allow robots to generalize better to unseen environments and settings zero-shot. One mechanism for this kind of reasoning would be world models, but traditional video world models still tend to struggle with long horizons, and are very data intensive to train. But what if instead of predicting images about the future, we predicted just the symbolic information necessary for reasoning? @nishanthkumar23 tells us about Pixels to Predicates, a method for symbol grounding which allows a VLM to plan sequences of robot skills to achieve unseen goals in previously unseen settings. To find out more, watch episode #44 of RoboPapers with @micoolcho and @chris_j_paxton now!

1:05:21