Unverified60% confidenceFactExact time
Certain AI capabilities are nearly zero when model scale is below a specific threshold, then rise sharply once the threshold is crossed, exhibiting a nonlinear 'phase transition' characteristic.
2
Sources
60%
Confidence
Long-term
Relevance
7/5/2026
First Seen
Sources
The AI Revolution: Facing Hard Problems Is a Privilege
twitterelder_plinius6/27/2026
Related Claims
UnverifiedThe binary threshold design is unfavorable to models near the threshold, as 0.74 and 0.80 differ drastically in score even though the actual capability gap may be negligible.72% similarUnverifiedAI生成假设的边际成本趋近于零，一旦模型训练完成每增加一条假设几乎不产生额外成本72% similarUnverifiedThe quality improvement brought by AI came from a decrease in the frequency of low scores and an increase in the frequency of moderate scores, while the proportion of excellent scores did not change, showing a floor-raising rather than ceiling-breaking effect.72% similarUnverified不同归一化方案对同一模型的IC影响可达0.01-0.03的量级71% similarUnverifiedWei等人在2022年系统记录了大模型涌现能力现象，发现某些能力在小模型上几乎为零而在大模型上急剧提升，呈现非连续的相变特征70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/112241API
curl https://kongchang.com/api/v1/knowledge/claims/112241MCP
get_claim(id=112241)