Fei-Fei Li

Assembled from 19 blocks · 2 evidence · 20 related

1. Story
2. The computer scientist who conceived and led ImageNet, on the argument — unfashionable when she made it — that the binding constraint on visual recognition was the absence of data at the scale of the real world rather than the design of the algorithms. The 2012 result vindicated that position more decisively than it vindicated any particular architecture. She is also a founding co-director of the Stanford Institute for Human-Centered Artificial Intelligence, which publishes the annual AI Index, the most widely used compilation of measurements about the field. The pairing is unusual and worth noting: the same person built the dataset that made benchmark-driven progress possible and now co-leads the institution that documents how far benchmark-driven progress can be trusted.
3. Knowledge
4. Benchmark saturation
5. Fei-Fei Li
6. Connections
7. ImageNet
8. Benchmark saturation
9. ImageNet is assembled and released
10. The AI Index records both fast benchmark movement and doubts about benchmarks
11. The difficulty of evaluating machine learning systems
12. Measured adoption of AI by firms
13. Fei-Fei Li
14. GLUE is saturated and SuperGLUE is built to replace it
15. An audit finds the most-cited preference leaderboard is optimised rather than merely measured
16. The AI Index records both fast benchmark movement and doubts about benchmarks
17. Evidence
18. An eight-layer convolutional network — five convolutional and three fully connected layers — trained on graphics processors with dropout regularisation achieved a top-5 error rate of 15.3% at ILSVRC 2012 against 26.2% for the second-placed entry, learning hierarchical features rather than using hand-designed ones. V55 verification basis: the paper was not fetched. The 15.3% and 26.2% figures and the architecture description were carried in multiple retrieved summaries and in a hosted copy of the paper itself, but the attribution in those summaries is to secondary explainers as often as to the paper, and a curator should confirm both error rates against the ILSVRC 2012 results table before publication.
19. AI benchmark scores rose rapidly across language, reasoning, coding and mathematics in 2025 while benchmarks themselves faced growing questions about reliability: resolution on SWE-bench Verified rose from about 60% to near 100% within a year; accuracy on OSWorld rose from roughly 12% to 66.3%; frontier models gained about thirty percentage points in one year on Humanity's Last Exam; and robots succeed at only about 12% of real household tasks. V55 verification basis: the report was not fetched, but these figures were returned by a search restricted to hai.stanford.edu and are attributed there to the Index itself. Model-by-model scores, leaderboard ratings and adoption percentages that appeared alongside them in retrieved synthesis were NOT separately confirmed and are excluded from this pack.