Unverified50% confidenceFactExact time
Algorithmic bias refers to unfair outputs that a model produces toward specific groups due to systematic flaws in training data or algorithm design, and academia typically divides it into two categories: data bias and representational bias.
1
Sources
50%
Confidence
Long-term
Relevance
10/7/2026
First Seen
Sources
Related Entities
Related Claims
Unverifiedexposure bias的根源在于训练与推理的分布不匹配，训练时使用teacher forcing策略而推理时模型依赖自身生成的token77% similarUnverifiedLLM-based AI personas may inherit biases from training data, including linguistic bias, socioeconomic bias, and temporal bias, leading to systematic misjudgment of minority groups or non-mainstream viewpoints.75% similarUnverifiedJudge models exhibit systematic biases: favoring output styles similar to their own training distribution, giving higher scores to longer responses (verbosity bias), and being sensitive to the position of answers in the prompt (position bias).74% similarUnverifiedWelfare approval algorithms may systematically reject applications from certain groups due to biases in training data, constituting institutional discrimination.74% similarUnverifiedThe direction of bias changes with the auditing format, indicating that bias is highly dependent on the tool's operationalization rather than being an intrinsic property of the model.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/985117API
curl https://kongchang.com/api/v1/knowledge/claims/985117MCP
get_claim(id=985117)