Unverified50% confidenceConceptExact time
Multiple large language models trained on large-scale web and book corpora (such as Claude, GPT, Gemini) have highly overlapping training distributions, so their consensus is correlated evidence rather than independent evidence and is insufficient to prove a conclusion.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
Unverified模型能力高度依赖训练数据的分布，法律文本充足而空间推理题的文本化表示相对稀少73% similarUnverifiedResearch shows that under certain conditions (such as a particular piece of text appearing at very high frequency in training data), models may reproduce verbatim content from training data, and in 2023 researchers from Google DeepMind and the University of California, Berkeley successfully extracted original fragments of training data from ChatGPT.69% similarUnverifiedMainstream benchmark problems have been widely discussed on the internet, and models' training data may contain memorization of these problems, causing benchmark scores to overestimate the models' capabilities on truly unseen tasks67% similarUnverifiedAdversarial multi-agent verification is good at detecting anchorable factual deviations (such as line numbers, timestamps, and hash values), but has limited effectiveness against systematic logical reasoning biases when sub-agents share training biases.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1017262API
curl https://kongchang.com/api/v1/knowledge/claims/1017262MCP
get_claim(id=1017262)