Unverified50% confidenceFactExact time
MERLIN's current experiments are concentrated on models at the 1.3B and 780M parameter levels, and performance on larger-scale models remains to be seen.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/28/2026
First Seen
Valid until: 1/26/2027
Sources
Related Entities
Related Claims
Unverified研究表明当模型参数规模达到3B以上时，1.58-bit模型在多项基准测试中表现可接近甚至匹配同等规模的全精度模型67% similarUnverified万象模型提供了 1.3B 和 14B 参数量的不同规模版本67% similarUnverifiedPhi-2 has 2.7B parameters and on multiple benchmark tests can match models several times its size.64% similarUnverifiedThe tests covered models in three price tiers: a local 1.2B routing small model, a local 8B-class backbone model, and a cloud-based primary model.63% similarUnverifiedFull-parameter fine-tuning typically requires GPU memory 3–5 times the model's parameter count.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1105935API
curl https://kongchang.com/api/v1/knowledge/claims/1105935MCP
get_claim(id=1105935)