Verified65% confidenceFactExact time
During model collapse, the model gradually loses the long-tail portions of the data distribution and ultimately converges to monotonous, repetitive, and diversity-lacking outputs.
3
Sources
65%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedIf only the data of newly added classes is used to fine-tune an expanded classification head, the model is highly prone to catastrophic forgetting, rapidly losing its ability to recognize the original classes.76% similarUnverified合成数据的反馈循环会导致模型崩溃，表现为词汇单一化、句子结构趋同、长尾内容消失75% similarUnverified错误的归纳偏置会让模型花费容量拟合数据中不存在的虚假模式，导致分布外数据上严重失效72% similarUnverifiedMulti-source data fusion without data governance produces a garbage-in-garbage-out problem, causing AI model outputs to lose credibility.71% similarUnverified用新数据更新模型时可能破坏原有能力，即灾难性遗忘，工业界通过失败案例库、数据增强与混合回放技术来整合新数据71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/900961API
curl https://kongchang.com/api/v1/knowledge/claims/900961MCP
get_claim(id=900961)