Curating the frontier of Artificial Intelligence

AutoMem: Automated Learning of Memory as a Cognitive Skill

记忆专业知识是一种习得的技能:知道要编码什么,何时检索,以及如何组织知识——这种能力在认知科学中被称为……

Memory expertise is a learned skill: knowing what to encode, when to retrieve, and how to organize knowledge--a capacity known in cognitive science as...

MIT in the media: Innovating and educating for the next 250 years of America

如果没有联邦政府对好奇心驱动的研究,的支持，帮助确保我们国家’繁荣和安全的创新和人才管道......

Without federal support for curiosity-driven research, the innovation and talent pipeline that has helped ensure our nation’s prosperity and safety co...

DOPD: Dual On-policy Distillation

在策略蒸馏 (OPD) 通过使用密集的令牌级信号监督学生采样轨迹，提供卓越的容量传输。提供...

On-policy distillation (OPD) offers superior capacity transfer by supervising student-sampled trajectories with dense token-level signals. To furnish ...

Self-Evolving World Models for LLM Agent Planning

世界模型提供了一种原则性的方法，使长期 LLM 代理在执行前对行动后果具有远见: 的预测。然而,不...

World models offer a principled way to equip long-horizon LLM agents with foresight: predictions of action consequences before execution. However, unr...

3 Questions: Beyond data

“Beyond 数据驱动美学,”，由麻省理工学院建筑系校友兼研究员 Alexandros Haridis, 在麻省理工学院凯勒画廊展出，直至 6 月 30, 展览...

“Beyond Data-Driven Aesthetics,” by MIT Architecture alumnus and researcher Alexandros Haridis, on view at the MIT Keller Gallery through June 30, exa...

Q&A: What is agentic AI today, and what do we want it to be?

Q: 什么是代理 AI，它与 ChatGPT 和 Claude 等生成型 AI 模型有何不同? A: 代理 AI 是在世界中采取行动的 AI。 ...

Q: What is agentic AI and how is it different from generative AI models like ChatGPT and Claude? A: Agentic AI is AI that takes actions in the world. ...

Inaugural Music Technology Research Showcase celebrates work of new graduate program’s initial students

这次庆祝活动以 MTC的前五名注册者 ( 为特色，他们都曾是麻省理工学院的本科生), 以及几名博士生和教职员工...

The celebratory occasion featured MTC’s first five enrollees (all of whom were previously MIT undergraduates), alongside several PhD students and facu...

Ask, Don't Judge: Binary Questions for Interpretable LLM Evaluation and Self-Improvement

评估 LLM 输出仍然是 NLP 的主要瓶颈: 人类评估成本昂贵且缓慢, 词汇指标与人类判断相关性很​​差......

Evaluating LLM outputs remains a major bottleneck in NLP: human evaluation is expensive and slow, lexical metrics correlate poorly with human judgment...

EO-WM: A Physically Informed World Model for Probabilistic Earth Observation Forecasting

地球观测(EO)预测旨在根据不断变化的气象条件下的卫星观测来预测未来地球表面的动态......

Earth Observation (EO) forecasting aims to predict future Earth surface dynamics from satellite observations under changing meteorological conditions....

Simulation-based inference for rapid Bayesian parameter estimation in epidemiological models: a comparison with MCMC

机制流行病学模型广泛用于支持传染病预测和公共卫生决策。贝叶斯校准...

Mechanistic epidemiological models are widely used to support infectious disease forecasting and public-health decision making. Bayesian calibration o...