ReactHuman: A Physics-Grounded Benchmark for Human-Like Reactive Decision-Making in Embodied Multimodal LLMs
- 类型：arxiv
- 标识：2609.10895
- 链接：https://arxiv.org/abs/2609.10895
- 主分类：multimodal
- 形态：benchmark
- 被引：1
- 被引来源：Semantic Scholar
- S2被引：1
- OpenAlex被引：0
- 影响力被引：0
- TLDR：Reacting to sudden physical hazards is both a meaningful test of embodied intelligence and a hard requirement for deploying multimodal large language models (MLLMs) as the decision core of household robots, and ReactHuman is the first physics-grounded benchmark for human-like reactive decision-making.
- OpenAlex ID：W7212300048
- OpenAlex DOI：10.48550/arxiv.2609.10895
- DOI：10.48550/arxiv.2609.10895
- DOI来源：OpenAlex
- 开放获取：green
- 开放获取链接：https://arxiv.org/pdf/2609.10895
- OpenAlex更新：2026-09-16
- 副分类：evaluation
- 待LLM分类：否
- 标题中文：ReactHuman: A Physics-Grounded Benchmark for Human-Like Reactive Decision-Making in Embodied Multimodal LLMs
- TLDR中文：对突发物理危险的反应既是具身智能的重要检验，也是将多模态大语言模型 (MLLMs) 部署为家庭机器人决策核心的硬性要求；ReactHuman 是首个面向类人反应式决策的物理驱动基准。
- 来源文件：
- /inbox/tom/_candidates/2026-09-15-rag-retrieval-reranking-candidates.json
- /inbox/tom/_candidates/2026-09-15-agent-rag-longcontext-candidates.json
- [OpenAlex backfill]
- [S2 enrich]