Unverified50% confidenceBenchmarkExact time
When models transitioned from historical texts to contemporary texts, a performance drop of 20% to 50% was observed.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/18/2026
First Seen
Valid until: 12/17/2026
Sources
Related Entities
Related Claims
UnverifiedMultiple course creators have reported income drops of more than 50%, reduced content engagement, and users turning to large language models.72% similarUnverified牛津大学2023年的一项研究量化了模型坍缩效应，仅经过五代自我训练，文本多样性指标下降了约40%70% similarUnverifiedOver the past year, token pricing for mainstream large language models has broadly dropped by more than 50%.70% similarUnverified实验数据显示，经过5-10轮迭代训练后，模型在低频词汇和长尾知识上的覆盖率下降超过40%，而输出文本在表面流畅性上几乎没有可察觉的退化69% similarUnverifiedResearch indicates that when the context exceeds a certain length, the model's recall accuracy for content in the middle portion may drop below 50%.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/931188API
curl https://kongchang.com/api/v1/knowledge/claims/931188MCP
get_claim(id=931188)