Unverified50% confidenceBenchmarkExact time
在评估阶段对候选关系标签进行随机打乱后，该模型的F1从37%骤降到14%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/5/2026
First Seen
Valid until: 11/3/2026
Sources
Related Claims
UnverifiedFor models with more than 30B parameters, the performance loss caused by 4-bit quantization typically results in a 1–3% drop on benchmarks, but the impact is more significant for smaller models.59% similarUnverified当相关文档被放置在上下文中间位置时，模型正确回答率相比放在首尾位置下降高达20-30个百分点59% similarUnverifiedThe lottery acceptance rate for the H-1B regular cap has fallen to approximately 15%–20% in recent years.59% similarUnverified量化推理从FP16降至FP8可将显存需求减半，从FP8降至FP4可进一步减半59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/692516API
curl https://kongchang.com/api/v1/knowledge/claims/692516MCP
get_claim(id=692516)