Zero-Shot Diffusion: Training-Free Image Generation Breakthrough
New training-free single-image diffusion models enable high-quality generation without retraining, slashing costs and la…
Latest articles in Research
New training-free single-image diffusion models enable high-quality generation without retraining, slashing costs and la…
Kuawei Intelligence tops the May 2026 WorldArena benchmark, signaling China's rising dominance in embodied world models.
Researchers find small models forget rare skills due to frequent data overwriting. Increasing target task frequency fixe…
New analysis reveals that Retrieval-Augmented Generation lacks clinical relevance without explicit persona modeling in h…
Researchers use generative AI to design a novel vaccine structure, marking a major leap in computational biology and dru…
ICRA 2026 concludes in Vienna with historic dual best paper awards and the launch of Tokyo University's new EVARL roboti…
Unlock the logic behind LLMs and AI agents with these 4 essential books on formal systems, philosophy, and reasoning.
Researchers at Xiaohongshu introduce Evolving-RL, a framework enabling autonomous skill evolution in AI agents without s…
AI tools are transforming scientific peer review, offering speed but raising concerns about bias and the erosion of huma…
New study reveals law faculty prefer AI-generated answers over peer submissions for grading tasks.
OpenAI's model autonomously solves the Erdős unit distance problem, marking a pivotal moment for AI in pure mathematics …
CMU researchers propose a 'sleep' mechanism for LLMs to consolidate long-context memory, solving KV cache bloat and impr…