Unverified50% confidenceFactExact time
作者观察横向覆盖 DeepSeek、GLM、Qwen 等多个开源模型均存在类似工具调用错误模式
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/22/2026
First Seen
Valid until: 10/20/2026
Sources
Related Claims
UnverifiedDeepSeek-R1、Qwen的QwQ等开源模型在积极探索类似o1的推理增强路线74% similarUnverifiedDeepSeek-R1和Qwen-QwQ等开源模型在积极探索类似o1的慢思考推理方向72% similarUnverifiedMeta的Llama系列、DeepSeek-Coder以及Qwen是开源阵营缩小与闭源模型代码能力差距的代表70% similarUnverifiedDeepSeek同步开源了名为DeepSpark的完整项目，包含训练、评估和实验代码，可移植到Qwen、Gemma等主流模型70% similarUnverifiedDeepSeek-R1和Qwen-QwQ等开源模型在推理增强（思维链推理）方向上取得了显著进展70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/593693API
curl https://kongchang.com/api/v1/knowledge/claims/593693MCP
get_claim(id=593693)