SGI-Bench

What is SGI-Bench?

SGI-Bench operationalizes this definition via four scientist-aligned task families: deep research, idea generation, dry/wet experiments, and multimodal…

Released

2025-12-22

Modality

multimodal

Evaluates

多模态, Multimodal, 科学, Science, 代码, Code, deep research, reasoning, lab protocol, 科学智能, AI for Science, 科学推理, Scientific Reasoning, 代码工程, 逻辑推理, 跨模态推理, Cross-modal Reasoning, 不支持, Unsupported

Openness

unknown

Importer

OpenCompass Hub

Reported scores

0

Source provenance

• Original evidence https://hub.opencompass.org.cn/dataset-detail/SGI-Bench

SGI-Bench paper and code

• Paper https://arxiv.org/pdf/2512.16969
• Code https://github.com/InternScience/SGI-Bench

Which sources cite SGI-Bench?

• SGI-Bench

Related 多模态, Multimodal, 科学, Science, 代码, Code, deep research, reasoning, lab protocol, 科学智能, AI for Science, 科学推理, Scientific Reasoning, 代码工程, 逻辑推理, 跨模态推理, Cross-modal Reasoning, 不支持, Unsupported benchmarks

• 3DCodeBench
• A-Bench
• A-OKVQA
• AA-AnalystAgent
• AA-Briefcase
• AA-LCR
• AA-LCR
• AA-Omniscience Index