Expired50% confidenceFactExact time
Gaokao essay scoring is conducted by cross-grading between two models, GPT-4.1 and Claude Sonnet 5, mimicking the gaokao dual-marking model.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/9/2026
First Seen
Valid until: 10/7/2026(expired)
Sources
Huawei Pangu 2.0 Flash on the Gaokao Benchmark: 613 Points, Stability Issues, and Cost Breakdown
bilibiliwycode_cn7/7/2026
Related Claims
VerifiedClaude Sonnet 4 and GPT-5 are classified as advanced models, suitable for handling complex reasoning tasks and architectural design.74% similarUnverifiedMainstream commercial models such as GPT-4o and Claude 3.5 Sonnet bill separately for input and output tokens.72% similarUnverifiedLLM-as-Judge范式使用能力更强的大语言模型（如GPT-4o或Claude 3.5 Sonnet）对被评估模型的输出进行打分72% similarUnverifiedThe evaluation adopted a hybrid approach using GPT-5.5 as the LLM judge alongside human taste scoring.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/330509API
curl https://kongchang.com/api/v1/knowledge/claims/330509MCP
get_claim(id=330509)