PEARL is at the Conference on Robot Learning (CoRL) 2026 in Austin, Texas, on 9-12 November, with five accepted papers, one oral presentation, two invited workshop talks, and live tactile demonstrations. Our work spans contact-rich manipulation, bimanual learning, visual representations, agentic reinforcement learning, and learning through large-scale simulation.
Papers Presentation schedule Workshops & demos Meet the team Programme poster
Five papers at CoRL 2026
MiTaS: Multi-Resolution Tactile Imitation Learning for Contact-Rich Robotic Manipulation
Tactile imitation learning for contact-rich manipulation, with live demonstrations at the conference.
BiCICLe: Bimanual Robot Manipulation via Multi-Agent In-Context Learning
Coordinating two robot arms through multi-agent in-context learning.
Robot-DIFT: Correspondence-Sensitive Diffusion Features for Contact-Rich Robot Manipulation
Diffusion-derived visual features for precise, contact-rich robot control.
HARBOR: A Harness Framework for Agentic Robot Reinforcement Learning
A framework for agentic workflows in robot reinforcement learning.
MolmoB0T: Large-Scale Simulation Enables Zero-Shot Manipulation
An Ai2 collaboration on learning manipulation through large-scale simulation, presented as both an oral and a poster.
Where to find the presentations
All times are local to Austin (CST, UTC−6). Conference activities take place at the JW Marriott Austin.
| Date & time | Presentation |
|---|---|
| Monday 9 November 11:30-12:00 | MolmoB0T, Oral session O3 |
| Monday 9 November 17:00-18:30 | BiCICLe, P1, poster #119, Griffin, 2nd floor HARBOR, P1, poster #106, Lone Star, 3rd floor MolmoB0T, P1, poster #196, Griffin, 2nd floor |
| Tuesday 10 November 17:00-18:30 | MiTaS, P2, poster #143, Griffin, 2nd floor |
| Wednesday 11 November 10:30-12:00 | Robot-DIFT, P3, poster #141, Griffin, 2nd floor |
Workshops, demonstrations, and service
Georgia Chalvatzaki is a CoRL 2026 Program Chair, alongside Ani Majumdar and Dieter Fox. On Thursday 12 November, she will give two invited workshop talks:
- 08:40, Structured Physical Intelligence (SPIN): “The Evolving Role of Structure in Physical Intelligence.”
- 10:00-10:25, Do Robots Need World Models?: invited talk, followed by participation in the 11:25-12:15 debate in the Brazos room. Times follow the workshop’s current published schedule.
PolyUMI at the UMI Arena: join Rickmer Krohn, Nils Dengler, and Georgia Chalvatzaki on Thursday 12 November from 15:30 for the workshop paper and device exhibition. Find poster P24, board JW3-14-2, and device D06. PolyUMI supports visual, tactile, and audio data collection for object inference and manipulation.
PEARL also contributes to LEAP organization, the R2RL breakout, Agentic Robotics: SUTO, and R2RL: In-Context Exploration.
Meet PEARL in Austin
Georgia Chalvatzaki, Nils Dengler, Vignesh Prasad, Christina Kassab, Franziska Herbert, Steven (Zechu) Li, Yufeng Jin, and Rickmer Krohn will be attending. Come by our posters, demonstrations, and workshops to discuss research and collaborations.
We are hiring two postdoctoral researchers
Our two openings focus on learning for control of dexterous, contact-rich manipulation and structured and interactive embodied intelligence. Read the position descriptions and apply
PEARL at CoRL: programme poster
Open or download the full-resolution poster
