Unverified50% confidenceFactExact time
The research team trained 849 models of different scales (four sizes in total) on 458,752 tasks to derive scaling laws.
1
Sources
50%
Confidence
Long-term
Relevance
10/28/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedHiPredict training used 81,717 training samples corresponding to 13,625 observations.66% similarUnverifiedMiniCPM5-2B open-sourced its training data, including approximately 500,000 agent samples and over 80,000 reinforcement learning samples.65% similarUnverified智谱GLM-5.3基于743B（7430亿参数）基座进行后训练，主打编程与智能体两大核心场景65% similarVerifiedLarge model training primarily consumes computing power from tens of thousands of high-end GPUs (such as NVIDIA A100/H100).65% similarUnverifiedThe OmniGround team built a 580K-sample training dataset by reconstructing temporal question answering from dense audio-visual annotations and organizing general omni-modal instruction data.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1106008API
curl https://kongchang.com/api/v1/knowledge/claims/1106008MCP
get_claim(id=1106008)