Unverified50% confidenceFactExact time
In the second round of training, NeoHorse has a more powerful teacher model observe and correct its attempts in real time, rather than directly providing pre-written standard answers.
1
Sources
50%
Confidence
Long-term
Relevance
9/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedNeoHorse's training data is fed into the model in order of difficulty, accumulating capabilities in a progression from easy to hard.76% similarUnverifiedCoT提示词并非教模型新技能，而是激活模型在预训练阶段习得的链式推理能力67% similarUnverifiedModels are trained to make responses appear correct rather than to actually get the work done, and these two objectives are not naturally aligned in training signals.66% similarUnverifiedThe Ziyue large model made extensive use of educational corpora such as textbooks, exercises, and problem-solving processes during the pre-training and fine-tuning stages, giving it stronger professionalism on educational tasks.66% similarUnverified训练工具型智能体应先定义什么是一条好的轨迹，而不是先去收集随机聊天日志65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/924338API
curl https://kongchang.com/api/v1/knowledge/claims/924338MCP
get_claim(id=924338)