Memento 3: Model-Based Recursive Self-Improvement through Reflective Rulebooks Paper • 2610.11794 • Published about 1 month ago • 37
Running 4 HOI-Retarget Contact Playback 🦾 4 Play back robot interaction clips with visual contact cues
iSDFT: Information-Proximal Self-Distillation for Continual Learning in LLMs Paper • 2609.24646 • Published Sep 21 • 9
iSDFT: Info-Proximal Self-Distillation Fine-Tuning Collection iSDFT Trained Models • 66 items • Updated Sep 22 • 4
From Skills to Talent: Organising Heterogeneous Agents as a Real-World Company Paper • 2604.22446 • Published Apr 24 • 102
Efficient and Principled Scientific Discovery through Bayesian Optimization: A Tutorial Paper • 2604.01328 • Published Apr 1 • 6
Efficient and Principled Scientific Discovery through Bayesian Optimization: A Tutorial Paper • 2604.01328 • Published Apr 1 • 6
Decoding as Optimisation on the Probability Simplex: From Top-K to Top-P (Nucleus) to Best-of-K Samplers Paper • 2602.18292 • Published Feb 20 • 14 • 7
The Y-Combinator for LLMs: Solving Long-Context Rot with λ-Calculus Paper • 2603.20105 • Published Mar 20 • 36
The Y-Combinator for LLMs: Solving Long-Context Rot with λ-Calculus Paper • 2603.20105 • Published Mar 20 • 36
Enhancing Spatial Understanding in Image Generation via Reward Modeling Paper • 2602.24233 • Published Feb 27 • 56
Recovered in Translation: Efficient Pipeline for Automated Translation of Benchmarks and Datasets Paper • 2602.22207 • Published Feb 25 • 45
CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation Paper • 2602.24286 • Published Feb 27 • 96
Mode Seeking meets Mean Seeking for Fast Long Video Generation Paper • 2602.24289 • Published Feb 27 • 40
Compositional Generalization Requires Linear, Orthogonal Representations in Vision Embedding Models Paper • 2602.24264 • Published Feb 27 • 14