Unverified50% confidenceFactExact time
The XiangqiBench research team constructed 119 tactical endgame positions with forced checkmates, verified through an engine or pure check search.
1
Sources
50%
Confidence
Long-term
Relevance
10/5/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe test ran a total of 130 variants, covering different levels of reasoning effort, called via OpenRouter with official endpoints for each lab locked in wherever possible.56% similarUnverifiedBenchLocal的测试维度包括工具调用、命令行操作、Bug识别修复、指令遵循、Hermes Agent、结构化输出、数学推理、数据提取共8大类55% similarUnverifiedAn independent researcher constructed 22 multi-turn traps and 84 single-turn traps, had 10 models run through all 106 tests, and independently scored each trajectory using three methods.55% similarUnverified成员推断攻击的基本思路源于Shokri等人2017年的工作，通过观察模型对特定样本的置信度或损失值判断该样本是否在训练集中54% similarUnverifiedXiangqiBench is an executable benchmark using Chinese chess as its vehicle, published on arXiv, and argues that agent evaluation should measure closed-loop outcomes and report both coverage and reliability.54% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/975455API
curl https://kongchang.com/api/v1/knowledge/claims/975455MCP
get_claim(id=975455)