Expired60% confidenceFactTime unknown
Exposure Bias is one of the root causes of AI hallucination: during training the model sees real text as context (Teacher Forcing), but during inference it must continue generating based on its own previously generated content, causing errors to accumulate and amplify.
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Related Claims
VerifiedAI幻觉的根源在于模型学习的是词与词之间的统计共现关系，在知识稀疏或矛盾区域倾向于生成听起来合理的内容而非承认不确定性78% similarVerifiedThe two root causes of the 'AI flavor' in AI-generated writing are training data bias and AI's lack of genuine creative intent.78% similarUnverifiedHaving the same Agent write code and then self-review it exhibits a problem analogous to human Confirmation Bias, which is a mapping of a cognitive psychology phenomenon onto AI systems.77% similarUnverifiedThere are vast differences in the ability of different AI models to understand implicit semantics and emotional expression, which are related to differences in training data composition, alignment strategies, and decoding strategies.75% similarUnverifiedAI编程带来的效率提升有时是主观感受而非客观事实，涉及流畅性启发式和自动化偏见等认知偏差75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/55871API
curl https://kongchang.com/api/v1/knowledge/claims/55871MCP
get_claim(id=55871)