Robin Walters

Asst. Prof. at Khoury College of CS at Northeastern

[Hannah Lawrence @HLawrenceCS]

@bostonsymmetry is hosting a poster session (4-5:30pm) + social on Tuesday, June 9th, at Northeastern! All are welcome to come chat about geometry, symmetries, + AI!

Location: Raytheon Amphitheater, https://t.co/uYjUlqwxzC

Register here: https://t.co/pR7FyySz8I

[TAGinDS @TAGinDS]

Just under two weeks until the submission deadline for the Boston TAG Party 2026– a joint conference organized collaboratively by the Boston Symmetry Group and TAG-DS! Papers are due June 12th— full archival papers, extended abstracts, and open problem tracks!

[Kostas Daniilidis @KostasPenn]

Last Thursday, @stefanos_pert Stefanos Pertigkiozoglou defended his PhD. Stefanos is one of the deepest and most thorough thinkers I worked with. He is the definition of quiet power. His key insight was that enforcing exact equivariance throughout optimization can unnecessarily restrict expressivity. By relaxing equivariance constraints during training, his methods reached solutions that preserve the benefits of symmetry while escaping the restricted equivariant landscapes (ICML'25 and Neurips'24, plus TMLR, ICLR, Neureps). I am so proud of you Stefanos!
Thank you @_onionesque for the inspiration and Stefanos' mentoring in the relaxation work, as well as to @RobinSFWalters, @ParisPerdikaris, @EdgarDobriban, Pratik Chaudhari, and Jean Gallier, for making committee exams a resource for learning.

[Lakshita Dodeja @ ICRA @lakshitadodeja]

Can BC policies be quickly improved through real world experience?

Our new #RSS2026 paper proposes Q2RL, a method that bridges BC and RL for on-robot learning.

Q2RL improves BC policies by up to 3.75x with just 1-2 hours of online interaction!

So when life gives you BC, make Q-functions! 🍋

Details in thread 🧵

[Robert M. Gower @gowerrobert]

And now we are very proud and humbled to have received the ICLR 2026 Honorable Mention award for this work https://t.co/RgsVc3iv0G Very fun to have found this useful math nugget that can actually speed-up LLM training.

[Hannah Lawrence @HLawrenceCS]

TL;DR: poster today at 3:15pm, P3-#1109!

Have you ever benchmarked your method on QM9, MD17, OC20, or ModelNet? It turns out that the 3D orientations of point clouds in commonly used datasets are highly non-random. How did we prove this and why should you care? 🧵

[Xupeng Zhu @XupengZ]

Introducing our recent work [ICLR2026] EquAct: An SE(3)-Equivariant Multi-Task Transformer for 3D Robotic Manipulation with Yu Qi, Yizhe Zhu, Robin Walters, and Robert Platt.

[Paper](https://t.co/jmPY0feDLw）
[Video](https://t.co/483oXjvxxf）
[Code](https://t.co/B2rD1BHGou

[Omkar Patil @op45_indian]

🚨New paper alert 🚨 from @rai_inst!
https://t.co/LWW4wbqhHt

🤖You robot policy is actually better than you think!

We find that for a given policy, ALWAYS denoising a single noise vector, which we call a ✨Golden Ticket ✨, leads to consistent performance improvements!

🧵...

[Marco Pacini @MarcoPacini4]

Our paper "On Universality of Deep Equivariant Networks" will appear at ICLR 2026.
We show that appropiate depth or readout layers are enough to reach universality up to the separation constraint imposed by the architecture, e.g., WL for IGNs.

https://t.co/SwHoTDBzYF

[GRaM Workshop at ICLR 2026 @GRaM_org_]

📢The second edition of ✨GRaM workshop✨ is here this time at #ICLR26.

🌟Submit your exciting works in Geometry-grounded representations.

We welcome submissions in multiple tracks i.e.
📄 Proceedings
📝extended abstract
👩‍🏫Tutorial/blogpost
as well as an exciting challenge!

[GRaM Workshop at ICLR 2026 @GRaM_org_]

📢 GRaM workshop is back!! This time at #ICLR2026
🌟Excited to be in Brazil 🇧🇷!

More information coming soon!

@iclr_conf @GRaM_org_

[Robin Walters @RobinSFWalters]

ISP can do equivariant learning from only RGB images. We leverage the spherical projection technique we previously used in Image2Sphere for prediction.

[Boce Hu @boce_hu]

Coauthors: @Dian_Wang_, @_davidklee, @HengT0812, @XupengZ, @HaojieHuang13, @RobotPlatt, @RobinSFWalters.

More details can be found in:
Website: https://t.co/e17B6hc19N
Paper: https://t.co/kWYJ9I0vSL
Video: https://t.co/zZYZ85kAch

[Boce Hu @boce_hu]

Coauthors: @Dian_Wang_, @_davidklee, @HengT0812, @XupengZ, @HaojieHuang13, @RobotPlatt, @RobinSFWalters.

More details can be found in:
Website: https://t.co/e17B6hc19N
Paper: https://t.co/kWYJ9I0vSL
Video: https://t.co/zZYZ85kAch

[Haibo Zhao @ZhaoHaibo47588]

Excited to share our #ICML2025 paper, Hierarchical Equivariant Policy via Frame Transfer. Our Frame Transfer interface imposes high-level decision as a coordinate frame change in the low-level, boosting sim performance by 20%+ and enabling complex manipulation with 30 demos.

[Bo Zhao @BoZhao__]

When and why are neural network solutions connected by low-loss paths?

In our #ICML2025 paper, we show that mode connectivity often arises from symmetries—transformations of parameters that leave the network’s output unchanged.

Paper: https://t.co/bZSO5foYfv
(1/6)

[Boston Symmetry Group @bostonsymmetry]

Registration is now open for Boston Symmetry Day on March 31! Sign up by March 21st at https://t.co/kjpiejs0gB

We have an exciting lineup of speakers (see our website: https://t.co/2V7RQwloCW ) 
Also featuring a poster session so you have a chance to present your awesome work!

[Boston Symmetry Group @bostonsymmetry]

Save the date -- Boston Symmetry Day 2025 will be held on March 31st, at Northeastern University!

Speakers and sponsors to be announced in the coming weeks, but you can expect another great lineup of talks, networking, and posters. We'll see you there!

[Bo Zhao @BoZhao__]

What can we learn from neural network model weights?

Join us for the Weight Space Learning Workshop at #ICLR2025! @iclr_conf

📄Accepting extended abstracts & full papers
🗓️Submission Deadline: Feb 4, 2025

🔗https://t.co/chEJappCS4

[Dian Wang @Dian_Wang_]

Honored to receive my first best paper nomination from @corl_conf! Had such a great time at #CoRL2024, huge thanks to the organizing committee and all my co-authors: Steve, David, @tarikkelestemur, @HaojieHuang13, @ZhaoHaibo47588, Mark, JW, @RobinSFWalters, @RobotPlatt!

[Haojie Huang @HaojieHuang13]

Generate the goal state and then infer the manipulation pick-place action. Feel free to check poster session 4 at #36 for details.