Expired50% confidenceBenchmarkExact time
The gap between all uncensored model variants and the base model on the GSM8K mathematical reasoning test is within 1.2 percentage points.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
9/7/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 9/21/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedThe optimal variant showed only a 1.83% decline on the GSM8K mathematical reasoning benchmark, while the MMLU general knowledge benchmark remained unchanged.73% similarVerifiedOn mainstream evaluations such as MMLU, HumanEval, and GSM8K, the score gaps among leading large models are often only a few percentage points.68% similarUnverified研究表明，即使在 GSM8K 或 MATH 数据集上接近满分的模型，面对稍作变形或需要跨领域组合的问题时性能下降可达30-50个百分点67% similarUnverifiedSelf-Consistency can bring improvements of 5–20 percentage points on mathematical reasoning benchmarks such as GSM8K.67% similarUnverifiedGLM5.2's inference speed is noticeably slower than models of the same tier, which is its biggest weakness.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/871881API
curl https://kongchang.com/api/v1/knowledge/claims/871881MCP
get_claim(id=871881)