2026/08 - Rome Focus Week Final Presentations

Video Discussions
1

The team presents the results from the Rome Focus Week Retreat. Team Mighty Mouse demos their ARC-AGI-3 integration, Team Fate Attenzione demos their new attention system prototype in Monty, and Team Janus demos a new ability to merge or split objects in Monty.

Main Video

0:00 Introductions
0:10 Rome Focus Week Recap
2:48 Team Mighty Mouse: ARC-AGI-3 (Dale, Jeremy, and Hojae)
4:45 ARC-AGI-3 Game Demo
7:04 Game Engine API for Custom Tasks
8:40 ARC-AGI Simulator
10:30 Sensor Module to Detect Changes
15:20 Learning Compositional Game Maps (and Sprites)
21:50 Mighty Mouse Conclusions
23:47 Mighty Mouse Q&A
31:04 Team Fate Attenzione: Building an Attention System in Monty (Scott, Viviane, Tristan)
44:21 Model-Free Attention Prototype
46:10 Model-Based Attention (Including Attention Inhibition)
48:29 Filtering Percepts
53:57 Adding a Background & Removing “Off Object” Filtering
58:39 Adding Soft Attention & No-Match Detection
1:01:05 Salience Change Detection & Explored Tracking Moving Objects
1:07:46 Dynamic Voxel Sizes
1:09:06 Fate Attenzione Summary & Q&A
1:14:50 Team Janus: Object Merging & Splitting (Ramy, Benji & Niels)
1:21:02 A New Web GUI Visualization
1:21:49 Web GUI Demo
1:23:13 Object Merging
1:31:57 Object Merging Experiments (Using Model Ablations)
1:37:51 Object Merging: Neuroscience Discussion
1:42:20 Object Splitting
1:44:27 Splitting Demo & Results
1:50:09 Object Splitting: Question about Bursting
1:54:10 Object Splitting Implementation Details
1:55:32 Future Work & Open Problems
1:57:31 Judges Interactive Demo: Monty in Browser
2:05:12 Closing & Judges Deliberation

1 Like
2

Incredibly interesting projects from all the teams yet again! :smiley:

About the ARC 3 test, even if it wasn’t meant to yield an immediate result, forcing to think differently is really the main benefit, as Hawkins mentioned in a past meeting. It’s also good preparation for the next ARCs. When you look at Astra’s replays, it’s poking things around and assigning semantic tags to the sprites based on their behavior and reactions, allowing the LLM to leverage all the pre-trained behavioral knowledge it has about those tags. It hints at abstract higher cognition being an initial requirement to properly recognize behaviors and affordances. (This doesn’t require semantic tags, but at the very least dimensionally-reduced prior knowledge of the involved concepts.)

The attention system is my personal favorite. I was initially a bit skeptical of the approach as the presentation started, but my few concerns were all addressed by the end of it. Very neat. I guess a more complex case to visit in the far future would be context-based attention, e.g. if I see an insect flying indoors from the corner of my eye, it will immediately grab my attention, but if I see one outdoors, I will likely ignore it. Well, unless it’s a wasp…

The web-based GUI is plain awesome. This is exactly the kind of stuff that can attract more people to the project. Great work.

Merging and splitting are well thought-out all around; quite clever to use sensorimotor exploration directly. Hawkins never stops throwing his insightful edge cases into the fray :laughing:. I guess some aspects of those edge cases would have to be addressed by abstract higher cognition, e.g. a unique object successfully keeping a consistent UUID over its lifetime even as its physical properties evolve (like losing a chip / being worn down).

1 Like