ModelsOfficial

SID AI 发布 SID-2 智能体搜索模型,Swarm 模式达 0.720 NDCG

SID AI Releases SID-2 Agentic Search Model With 0.720 NDCG Swarm Mode

Why we picked it

SID AI 把搜索做成了智能体:单个 3.11 秒拿 0.585 NDCG,12 个一起跑冲到 0.720,还压过 GPT-6 Astra。不过成绩是他们自家基准测的,看看多智能体并行搜索这个思路就好。

SID AI 发布了用强化学习训练的智能体搜索模型 SID-2。单智能体在 3.11 秒内取得 0.585 NDCG;12 个智能体组成的 SID-2 Swarm 达到 0.720 NDCG,超过 GPT-6 Astra。

Image · marktechpost
Body

技术机制上,SID-2 通过强化学习训练执行搜索任务,并支持多智能体协作的 Swarm 模式,让多个 SID-2 实例并行检索后汇总结果,以此把成绩从单体的 0.585 提升到 12 智能体的 0.720,代价是时延从 3.11 秒增加。

需要注意,0.720 NDCG 是在 SID 自家基准上测得,对比对象 GPT-6 Astra 也在同一基准下评估,与其他公开基准没有直接可比性。对做检索增强(RAG)或搜索产品的开发者来说,多智能体换精度的思路值得对照自家场景权衡时延成本。

Original · marktechpost

SID AI Releases SID-2 Agentic Search Model With 0.720 NDCG Swarm Mode

SID AI released SID-2, an agentic search model trained with reinforcement learning. A single agent reaches 0.585 NDCG in 3.11 seconds, and a 12-agent SID-2 Swarm scores 0.720 NDCG, above GPT-6 Astra on SID’s benchmark. The post SID AI Releases SID-2 Agentic Search Model With 0.720 NDCG Swarm Mode appeared first on MarkTechPost .

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Models" desk. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include SID-2, SID AI, GPT-6 Astra, agents.

Primary source verification: published directly by marktechpost with first-party authority. Cross-source corroboration: 11 independent sources have covered this development (including arXiv cs.AI, Decoder, Ate-a-Pi), confirming industry-wide signal strength.