Unverified50% confidenceConceptExact time
评测游戏（Benchmark Gaming）指AI公司通过选择性使用特定基准测试来最大化模型表现分数的行为，包括数据污染和针对特定评测集过度优化
1
Sources
50%
Confidence
Long-term
Relevance
7/14/2026
First Seen
Sources
Zig作者批评Anthropic：AI宣传与技术现实的落差
Related Claims
UnverifiedFrom checkpoint testing to official release, a model typically still needs to go through processes such as fine-tuning optimization, safety hardening, inference efficiency optimization, and API adaptation.67% similarUnverifiedBenchmark hacking refers to models potentially being optimized for specific test sets rather than possessing general capabilities, resulting in a gap between benchmark scores and real-world performance.66% similarUnverifiedBenchmark hacking can artificially inflate benchmark scores through methods such as data contamination (test set leakage into training data), targeted training on the specific question formats of particular benchmarks, and selective reporting of the best benchmark results.65% similarUnverifiedBenchmarks tend to evaluate how much a model knows, rather than how much cognitive increment it can bring.65% similarUnverifiedAI evaluation methods currently being explored by the industry include human-annotation-based scoring systems, BLEU scores, ROUGE scores, and the LLM-as-Judge method.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/508068API
curl https://kongchang.com/api/v1/knowledge/claims/508068MCP
get_claim(id=508068)