Unverified50% confidenceFactExact time
Since 2024, further simplified alignment algorithms such as ORPO, SimPO, and KTO have emerged.
1
Sources
50%
Confidence
Long-term
Relevance
8/8/2026
First Seen
Sources
GPT-5.6 Sol持续优化，Luna向免费用户开放：模型分层策略解读
Related Claims
Partially VerifiedO4 Mini is a reasoning-optimized model launched by OpenAI in 2025 and belongs to the O series.65% similarVerifiedContinuous Batching (iteration-level scheduling) was first proposed by the Orca system, published in an OSDI 2022 paper.64% similarUnverifiedThe concept of iteration-level scheduling was first proposed by the Orca paper published in 2022.64% similarUnverifiedGRPO emerged in 2024 and was adopted by DeepSeek-R1, further simplifying the reference model dependency on top of DPO63% similarUnverifiedFrom ChatGPT in 2022 to the o1/o3 series in 2024, the reasoning capabilities of large language models have undergone a qualitative leap.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/709644API
curl https://kongchang.com/api/v1/knowledge/claims/709644MCP
get_claim(id=709644)