Unverified50% confidenceCautionExact time
On the GSM8K mathematical reasoning task, Qwen model adaptation can maintain answer format compliance, but the exact-match retention effect is confounded by random seeds at the 0.6B scale and even deteriorates at the 1.7B scale.
1
Sources
50%
Confidence
Long-term
Relevance
10/7/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAfter models approached perfect scores on GSM8K, researchers immediately needed to introduce harder benchmarks, giving rise to the phenomenon of benchmark saturation.67% similarUnverifiedWithout changing the prompt, the capabilities of the SymCE-trained model transfer to mainstream mathematical reasoning benchmarks such as GSM8K, MATH-500, and MMLU-college-math.66% similarUnverifiedReverse-engineered prompt results may not 100% reproduce the original image, because the original prompt is highly dependent on a specific model version and random seed.65% similarUnverified2022年Wei等人的论文《Chain-of-Thought Prompting Elicits Reasoning in Large Language Models》证明在few-shot示例中加入推理步骤可将PaLM 540B在GSM8K上的准确率从17.9%提升至58.1%65% similarUnverifiedThe smaller model size of 7B parameters gives Qwen Image 2.1 advantages in local deployment and inference costs.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/982983API
curl https://kongchang.com/api/v1/knowledge/claims/982983MCP
get_claim(id=982983)