Unverified50% confidenceFactExact time
GSM8K, MATH, and BBH are three commonly used reasoning benchmarks with a clear difficulty gradient: GSM8K consists of elementary school math word problems, MATH covers competition-level math problems, and BBH contains 23 task types requiring multi-step symbolic manipulation or strategic reasoning.
1
Sources
50%
Confidence
Long-term
Relevance
10/20/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedGSM8K是小学数学推理基准，HumanEval是代码生成基准，HellaSwag是常识推理基准75% similarUnverifiedThere is a qualitative gap between the currently popular evaluation benchmarks GSM8K and MATH and frontier research-level mathematical reasoning72% 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.64% 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%60% similarUnverifiedThe inference bottlenecks of MoE models lie mainly in two areas: KV Cache read/write (constrained by memory bandwidth) and GEMM computation for MoE-routed experts (a compute-intensive task with irregular shapes).60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1073974API
curl https://kongchang.com/api/v1/knowledge/claims/1073974MCP
get_claim(id=1073974)