GPT-4o能拼好乐高吗？首个多步空间推理评测基准来了：闭源模型领跑，但仍远不及人类
"空间推理" 相关文章
GPT-4o能拼好乐高吗？首个多步空间推理评测基准来了：闭源模型领跑，但仍远不及人类
#AAAI2025 invited talk round-up 1: labour economics, and reasoning about spatial information
大模型不再是路痴！空间推理的答案是RAG：旅游规划、附近推荐全解锁
2025.03.27 | Dita跨模态策略优异，Qwen2.5-Omni多模态实时响应。
Google DeepMind’s Gemini Robotics: Unleashing Embodied AI with Zero-Shot Control and Enhanced Spatial Reasoning
This AI Paper from Menlo Research Introduces AlphaMaze: A Two-Stage Training Framework for Enhancing Spatial Reasoning in Large Language Models
超过o1-mini、逼近o3-mini，DeepSeek-R1在新的贪吃蛇基准上拿下1801分
超越CoT！微软剑桥中科院提出MVoT，直接可视化多模态推理过程
AAAI 2025 | 多模态大语言模型空间智能新探索：仅需单张图片或一句话，就可以精准生成3D建模代码啦！
「空间推理」成大厂竞逐焦点，为什么让大模型理解「内外远近」更重要？
李飞飞谢赛宁：多模态LLM「空间大脑」觉醒，惊现世界模型雏形！
谢赛宁、李飞飞「空间智能」新作：多模态大模型仍难突破空间推理
This AI Paper Proposes a Novel Neural-Symbolic Framework that Enhances LLMs’ Spatial Reasoning Abilities
Researchers at Stanford University Propose Locality Alignment: A New Post-Training Stage for Vision Transformers ViTs
A Visual Task that's Hard for GPT-4o, but Doable for Primary Schoolers
