LFM2-ColBERT-350M is also very fast! Its inference speed is on par with GTE-ModernColBERT-v1 (only 150M parameters) for query and document encoding across various batch sizes.
MACHINE LEARNING
-
Autonomous Cleaning Robot: Self-Learning Precision Intelligence
By
–
Un simple descuido puede parecer un desastre… pero no para él.
— Juan Merodio (@juanmerodio) 28 octobre 2025
Este robot detecta, actúa y se perfecciona solo, volviendo a su base para seguir dejando cada rincón impecable.
Porque el futuro no se imagina, se limpia con precisión, inteligencia y elegancia. pic.twitter.com/guh1gnR8XmUn simple descuido puede parecer un desastre… pero no para él. Este robot detecta, actúa y se perfecciona solo, volviendo a su base para seguir dejando cada rincón impecable. Porque el futuro no se imagina, se limpia con precisión, inteligencia y elegancia.
-
Technical analysis of data and architectural modules in generative environment modeling
By
–
Shows what happens when you remove key modules. – Without data curation → objects intersect walls
– Without coarse-to-fine → chaotic layouts With both → structured, realistic environments This figure is the mic drop infinite worlds need structured thinking. -
WorldGrow: A New AI System for Generative 3D City Modeling
By
–
Not just houses.
WorldGrow generates cities. Urban streets, suburban neighborhoods, consistent architectural styles all trained from UrbanScene3D. It’s the first system that can generate both indoor and outdoor 3D worlds without retraining. -
Breakthrough in Infinite 3D World Generation Using AI
By
–
The big one. 19×39 blocks ~1,800 m² of continuous 3D environment. Left: reconstructed mesh.
Right: fully textured render. It proves you can scale world generation infinitely without collapse or drift. -
WorldGrow Outperforms Generative Models in Continuous Indoor Scene Synthesis
By
–
Here’s where it flexes against SynCity, BlockFusion, DiffInDScene, and TRELLIS. Only WorldGrow generates high-res, continuous indoor scenes with perfect texture alignment. Figure 5 literally shows you the other methods fall apart at the seams.
-
WorldGrow improves 3D modeling via occlusion-aware feature aggregation
By
–
Standard 3D models fail near occlusions and boundaries. WorldGrow fixes this with occlusion-aware feature aggregation and a retrained decoder removing color bleeding and floating geometry. Figure 3 visually compares the “vanilla” vs “scene-friendly” SLAT. The difference is
-
Technical Overview of WorldGrow’s Generative AI Pipeline
By
–
WorldGrow’s pipeline works like a growing brain: – Scene-friendly SLAT encodes 3D context
– 3D block inpainting ensures spatial continuity
– Coarse-to-fine refinement keeps global layout + fine detail Each module adds realism while keeping the world endless. -
Volume 287 ICMLIC Proceedings Published on PMLR
By
–
Volume 287 https://
proceedings.mlr.press/v278/ Proceedings of International Conference on Machine Learning and Intelligent Computing Is now available on PMLR -
Training a GPT-2 Style Tokenizer and Model
By
–
Tu parles de tokenizer j’entraîne un tokenizer sur mon model. GPT2 est un modèle décodeurs only c’est que je vais. J’entraîne dans l’esprit un meme model que gpt2. Pour entraîner gpt2 ils faut plusieurs dizaines de milliers de dollars parce qu’il faut gérer des centaines de