Unverified50% confidenceFactExact time
训练数据归因（TDA）的常见技术路径包括影响函数、相似度检索和成员推断攻击
1
Sources
50%
Confidence
Long-term
Relevance
7/13/2026
First Seen
Sources
纽约时报指控OpenAI隐匿证据，ChatGPT版权诉讼再升级
rss7/9/2026
Related Claims
UnverifiedAdversarial training improves a model's ability to recognize and resist attack prompts by incorporating known jailbreak samples into training data, but faces the ongoing challenge of novel variants.71% similarUnverifiedThe training recipes published by SWE-bench cover key aspects including data preparation, model selection, fine-tuning strategies, and inference frameworks.67% similarUnverified蒸馏攻击通过大规模API调用收集输入-输出对，训练效率可能是常规预训练数据的数十倍64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/497638API
curl https://kongchang.com/api/v1/knowledge/claims/497638MCP
get_claim(id=497638)