Unverified60% confidenceBenchmarkExact time
Experiments conducted by the research team on models such as GPT-2 showed that after only 5–9 generations of recursive training, the quality of model outputs degrades significantly.
2
Sources
60%
Confidence
Long-term
Relevance
8/11/2026
First Seen
Sources
AI伪造创造力：当每间办公室的创意都在趋同
Related Claims
UnverifiedIn 2023, researchers from Oxford University and other institutions published a study finding that after only 5–10 generations of recursive training, the degradation in model output quality became visible to the naked eye.74% similarUnverified即使只有5-10%的训练样本存在错位，也会导致VLM模型在下游任务上出现显著的性能退化73% similarUnverifiedLoRA reports on GPT-3 175B: compared with full fine-tuning using Adam, trainable parameters are reduced by about 10,000 times, and GPU memory during training is reduced by about 3 times.72% similarUnverifiedIn 2023, researchers demonstrated that contaminating less than 1% of training data is sufficient to alter model behavior under specific trigger conditions72% similarUnverifiedWhen sampling eight candidates without constraints from the model, 10.2% of inputs cannot produce any valid output under limited training data, and this proportion drops to 0.4% when using the full training data71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/729465API
curl https://kongchang.com/api/v1/knowledge/claims/729465MCP
get_claim(id=729465)