WorldStereo is a method that turns a single photo (or a panorama) into a short set of camera-guided videos and then reconstructs a consistent 3D scene from them.
MMR-Life is a new test (benchmark) that checks how AI understands everyday situations using several real photos at once.
CharacterFlywheel is a step‑by‑step loop that steadily improves chatty AI characters by learning from real conversations on Instagram, WhatsApp, and Messenger.
CoVe is a way to create training conversations for AI agents that use tools, while guaranteeing the conversations are both challenging and correct.
Reinforcement learning (RL) trains language models by letting them try answers and learn from rewards, but training is slow if we pick the wrong practice questions.
The paper teaches AI agents to understand big codebases without running the code by following a strict, step-by-step thinking template called semi-formal reasoning.
FireRed-OCR turns a general vision-language model into a careful document reader that follows strict rules, so its outputs are usable in the real world.
OpenAutoNLU is a simple, open-source tool that automatically builds text understanding models for you.
Legal RAG Bench is a new, end-to-end test that checks how well legal AI systems find information and use it to answer tough, real-world legal questions.
The paper introduces SPOT, a training recipe that fixes an AI model’s mistakes with tiny edits while keeping what it already knows well.
Longer explanations are not always better; the shape of thinking matters.
RubricBench is a new benchmark that checks whether AI judges can use clear, checklist-style rules (rubrics) the way humans do.