Instruction-guided image editors have become highly capable, yet improving them further st…
While learning dexterous manipulation from a single human video offers a promising alterna…
We present DreamTrue, a multi-view, cross-embodiment robot world model for action-faithful…
Text-conditioned motion generators produce trackable whole-body motion, but they have no n…
General-purpose robots must perform a wide range of tasks from agile locomotion to dextero…
World models offer a promising alternative to physics-based simulators, yet remain far fro…
In 2026, cybersecurity evaluations involving OpenAI, Anthropic, and Google agents reached …
Recent 3D world models generate photorealistic, explorable scenes that remain frozen in ti…
Video generators and video-based world models can synthesize plausible visual trajectories…
Multimodal large language models (MLLMs) are rapidly evolving toward continuous audio--vis…
On a film set, the camera move is committed during a take. Generative video reshooting let…
Vision-Language-Action (VLA) models have emerged as powerful foundations for robotic manip…
Bifurcations are ubiquitous in physical systems, from structural buckling to fluid and cli…
In this work, we show that a single Transformer block, applied recurrently, can match the …
Recent incidents have highlighted the challenge of monitoring LLM agents and the danger of…
Quantizing AdamW's optimizer states reduces persistent storage, but quantization errors pr…
Generating an egocentric video from a single exocentric recording is a challenging case of…
Density ratios quantify distribution shift from a probability-mass point of view, whereas …
Teaching a humanoid to follow instructions with its whole body runs into two obstacles. It…
Safe reinforcement learning commonly places safety and task performance in the same policy…
Building a quantum model involves a tradeoff: how complex the circuit should be, and how m…
孙宇晨 × 邵艾伦 4 小时对谈：全文 + 精华 + 一键丢给 AI 的规划 prompt
跨 AI 桌宠：跟着你用的 AI 换形象、跟着 AI 的工作状态做动作
Ambi is releasing apps for iPhone, Mac, and Apple Watch that enables ambient capture
游戏打到21:0，AI甚至不用再调用大模型了
GLM「速度仙人」来了？
On Equity, we debate the impact of data center moratoriums and whether the AI industry has…
YouTube · Two Minute Papers