Shweta Mahajan

Assistant Professor at York University || Vector Institute
Postdoc, University of British Columbia || Ph.D., TU Darmstadt

[Rajeev Yasarla @RajeevYasarla]

Our ICML 2026 paper, RoCA: Robust Cross-Domain End-to-End Autonomous Driving!!

Paper link: https://t.co/Foiv0BbkNA
#ICML2026 #AutonomousDriving #AI #MachineLearning #ComputerVision #Robotics #SelfDrivingCars

[Shweta Mahajan @Matewhs]

@CSProfKGD @queensu Congratulations! 👏

[Apratim Bhattacharyya @apratimbh]

🚨 The project page is now live: https://t.co/jn0304hjps (Full code coming soon!)

[Shweta Mahajan @Matewhs]

@CSProfKGD @jbhuang0604 Wow!! Amazing!

[Shweta Mahajan @Matewhs]

Our new benchmark and dataset for interactive agents. The AI assistants are expected to interrupt when a mistake is made and provide corrective feedback! The benchmark is high-quality collected by our own research team! @apratimbh @RolandMemisevic

[Apratim Bhattacharyya @apratimbh]

🚨Introducing: Ego-MC-Bench (Mistake Corrections) benchmark and Ego-CoMist (Counterfactual Mistakes) dataset.

🎯Ego-MC-Bench: Where AI assistants need to intervene at the right time (when) and with the right feedback (what) to prevent mistakes.

👉https://t.co/zFNEZWAMpt

1/4

[Apratim Bhattacharyya @apratimbh]

🚨Introducing: Ego-MC-Bench (Mistake Corrections) benchmark and Ego-CoMist (Counterfactual Mistakes) dataset.

🎯Ego-MC-Bench: Where AI assistants need to intervene at the right time (when) and with the right feedback (what) to prevent mistakes.

👉https://t.co/zFNEZWAMpt

1/4

[Shweta Mahajan @Matewhs]

@anand_bhattad Congratulations!

[Shweta Mahajan @Matewhs]

@PrincetonCS @adjiboussodieng @NSF congratulations @adjiboussodieng !

[Sahithya Ravi @Sahithya_Ravi]

The next talk will be Trevor Darell @trevordarrell !

Join us at rooms 610/612!

[Sahithya Ravi @Sahithya_Ravi]

Happening at 1 PM!
Come join us!!!!!

#cvpr2026

[Shweta Mahajan @Matewhs]

Vicente Ordóñez-Román presenting in 20 min! Hear from the OGs of Vision and Language research! Room 102 #CVPR2026 @CVPR

[Apratim Bhattacharyya @apratimbh]

📢Exciting schedule for VAR Workshop @CVPR today:

08:45-09:15: @ryoo_michael
09:15-09:45: Katerina Fragkiadaki
09:45-10:15: @WenhuChen
10:15-10:30: Molmo2 by @zixianma02
10:30-11:00: @liuziwei7
11:00-11:30: @YaoQin_UCSB
11:30-12:00: Vicente Ordóñez-Román

📍Room 102/104

[Apratim Bhattacharyya @apratimbh]

Happening now: Michael Ryoo

[Adi Chinchure @adityachinchure]

Today’s the day! Come attend our workshop on cognitive-inspired VLMs, starting at 1 PM

#CVPR2026

[Roland Memisevic @RolandMemisevic]

Happening this morning (Jun 3rd) in room 102!

[Shweta Mahajan @Matewhs]

Happening this morning in room 10
2!

[Apratim Bhattacharyya @apratimbh]

📢Exciting schedule for VAR Workshop @CVPR today:

08:45-09:15: @ryoo_michael
09:15-09:45: Katerina Fragkiadaki
09:45-10:15: @WenhuChen
10:15-10:30: Molmo2 by @zixianma02
10:30-11:00: @liuziwei7
11:00-11:30: @YaoQin_UCSB
11:30-12:00: Vicente Ordóñez-Román

📍Room 102/104

[Shweta Mahajan @Matewhs]

@YangWilliam_ Interesting! We did similar work by leveraging context for caption diversity in our NeurIPS 2020 work: https://t.co/koQnjbrOAK

[Sanjay Haresh @SanjayHaresh]

📢Excited to be presenting our work on memory + VLAs at ICRA'26 this Thursday morning (poster 224).

We found that a super simple language-based scratchpad with spatial and temporal grounding goes a long way in imparting memory to VLAs.

1/n

[Apratim Bhattacharyya @apratimbh]

📢Current VLAs are stateless - this is not ideal for long horizon tasks - which requires models to remember key past information, e.g., positions, orientations, task progress.

🎯So we just let the VLA take notes to help remember the past a.k.a "Notes-to-Self"!

👉ICRA 2026

[CogVL @ CVPR 2026 @CogVLWorkshop]

CogVL Workshop at #CVPR2026 is less than a week away!

We have an exciting lineup of keynote speakers across Vision, NLP and Cog Sci, and orals/posters on reasoning methods for VLM models.

🕐 June 3, 1 PM
📍 Rooms 610/612
🔗 Schedule: https://t.co/kNSQmKZuGJ

[Kosta Derpanis (sabbatical in Zurich) @CSProfKGD]

Attending #CVPR2026, Denver?

We have a stellar speaker lineup for “GeoFreeNVS: Geometry-Free Novel View Synthesis and Controllable Video Models.”

Come early. We anticipate seats will fill up FAST!

1/2