Daily Papers of 2025-12-26

1. Latent Implicit Visual Reasoning 63 upvotes, #1 of 2025-12-26
2. Emergent temporal abstractions in autoregressive models enable hierarchical reinforcement learning 59 upvotes, #2 of 2025-12-26
3. Spatia: Video Generation with Updatable Spatial Memory 28 upvotes, #3 of 2025-12-26
4. Schoenfeld's Anatomy of Mathematical Reasoning by Language Models 14 upvotes, #4 of 2025-12-26
5. VA-π: Variational Policy Alignment for Pixel-Aware Autoregressive Generation 10 upvotes, #5 of 2025-12-26
6. How Much 3D Do Video Foundation Models Encode? 9 upvotes, #6 of 2025-12-26
7. GTR-Turbo: Merged Checkpoint is Secretly a Free Teacher for Agentic VLM Training 3 upvotes, #7 of 2025-12-26

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.