Unverified50% confidenceFactExact time
Geirhos等人于2020年在Nature Machine Intelligence上发表综述，梳理了神经网络倾向于学习最简单判别规则而非人类期望语义特征的现象
1
Sources
50%
Confidence
Long-term
Relevance
8/10/2026
First Seen
Sources
1500组物体形变配对图像数据集：损伤检测AI训练数据的价值解析
Related Claims
Unverified2020年OpenAI的Kaplan等人论文首次系统证明神经网络性能与模型参数量、训练数据量、计算量之间存在幂律关系66% similarUnverifiedThe MIT 6.390 lecture notes cover a complete knowledge chain from the fundamentals of supervised learning (linear classifiers, perceptrons, SVMs) to neural networks and an introduction to reinforcement learning, with exercises accompanying each chapter65% similarUnverifiedThe systematic application of the chain rule to neural network training was not widely recognized in the deep learning field until Rumelhart, Hinton, and Williams' 1986 paper.63% similarUnverifiedMIT 6.390 'Introduction to Machine Learning,' formerly numbered 6.036, is taught by professors including Leslie Kaelbling and Tomás Lozano-Pérez, and its lecture notes cover complete content from linear classifiers and neural networks to reinforcement learning.62% similarUnverified该自进化架构借鉴了元学习（Meta-Learning）和神经架构搜索（NAS）领域的核心思想62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/723116API
curl https://kongchang.com/api/v1/knowledge/claims/723116MCP
get_claim(id=723116)