Tag: #academia17 results
Su Hao, author of ImageNet and Fei-Fei Li's top student, returns to China to teach at Fudan University. As the most cited researcher in embodied AI, he is appointed Dean of the General Physical AI Institute. This role focuses on advancing physical intelligence.
This article explores the historical and academic influence of Tsinghua University in the development of AI in China, highlighting its continuous contribution to the field over 48 years.
The article argues that AI has commoditized the 'execution' phase of research, necessitating a shift toward 'Registered Reports' that prioritize high-quality questioning and logic over mere data output. It suggests that academic institutions must evolve their evaluation metrics to value honesty and problem-solving over traditional publication counts.
A discussion regarding the realistic timeline for developing and publishing research papers at top-tier conferences like ICML, NeurIPS, and ICLR. It explores the lifecycle from initial idea to final acceptance.
National Institute of Informatics (NII) develops open LLMs with high Japanese performance. Director Kurohashi explains why academia pursues domestic models, stressing transparency as key to trust and reliability.
This discussion explores whether the integration of LLMs into research workflows, such as experiment execution and paper writing, is effectively shortening the duration of CS PhD programs. It questions if current academic bottlenecks are being bypassed by AI-assisted productivity.
探討為何 ICML 和 NeurIPS 等頂級會議在機器學習領域的聲望已超越傳統學術期刊。討論重點在於 AI 領域的快速發展需求以及會議審稿流程的效率優勢。
Post critiques modern ML PhDs as often incremental: combining ideas, tweaking for benchmarks, claiming SOTA without deeper insights. Questions if top papers yield lasting knowledge and ponders field incentives. Asks what distinguishes strong incremental work.
MIT is exploring ways to simplify startup creation for faculty and students. This aims to capitalize on surging AI demand. The initiative seeks to unleash innovation from campus.
Reddit post claims industry dominates ML research with superior compute and talent, leaving academia to niche, unrealistic, or outdated topics. Examples include GAN analysis, white-box attacks, ML misapplications, and deprecated model surveys. Notes researchers shifting to industry or startups.