Unverified50% confidenceFactExact time
Smaller models perform more poorly on script-following tasks.
1
Sources
50%
Confidence
Long-term
Relevance
9/25/2026
First Seen
Sources
Related Entities
Related Claims
Unverified研究表明模型在处理约束性指令时的准确率显著低于处理生成性指令，这被称为'约束遵循缺陷'74% similarVerifiedSmaller models often show exponential rather than linear degradation in performance on long reasoning chain tasks.72% similarVerifiedSmaller models are more prone to embedding collapse than larger models, due to reasons including limited parameter capacity, a tendency toward optimization shortcuts, and amplified anisotropy issues.71% similarUnverifiedChoosing a small causal language model reduces computational cost and makes experimental phenomena easier to observe, since larger models may exhibit entirely different resistance to small-scale fabricated corpora.71% similarUnverifiedLess capable local small models sometimes generate tool-call instructions with non-compliant formatting, causing operations to fail.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/949174API
curl https://kongchang.com/api/v1/knowledge/claims/949174MCP
get_claim(id=949174)