GPT-Image-2 Example · ReAct Reasoning Trajectory

A reference case from the "Research Paper Illustrations" selection of GPT-Image-2, suitable for exploring generation effects in the "ReAct Reasoning Trajectory" direction.

Author: AI Plus Lab

✦Results

◌Case Background

来自 GPT-Image2-Skill README 的精选展示条目，适合作为“研究论文图示 / ReAct 推理轨迹”方向的站内参考案例。

⌘Prompt Content

Horizontal 16:9 illustration, ReAct trajectory, for factual question-answering tasks, vertically arranged in 7 alternating blocks.

Top title: "Task — user asks: 'What year did the scientist who proved the Higgs boson exists win the Nobel Prize?'".

7 blocks from top to bottom, numbered 1–7 on the left side:
1. Thought: "I need to identify the scientist associated with the proof of the Higgs boson and then look up their Nobel Prize year."
2. Action: wiki_search("Higgs boson discovery")
3. Observation: "The 2012 announcement at CERN confirmed the Higgs boson..."
4. Thought: "The theoretical prediction is due to Peter Higgs and François Englert. I should check if they were later awarded the Nobel."
5. Action: wiki_search("Peter Higgs Nobel Prize")
6. Observation: "Peter Higgs and François Englert won the 2013 Nobel Prize in Physics..."
7. Thought: "Answer: 2013."

Thought blocks: left-side dust turquoise border, italicized, with brain icon. Action blocks: left-side soft navy blue border, monospaced font, with wrench icon. Observation blocks: left-side soft terra-cotta border, lighter fill, with eye icon. Thin slate gray arrows between blocks.

Bottom: capsule shape, "Final answer: 2013," with a checkmark icon.

Title: "ReAct trace: interleaved reasoning and tool-use on a factual-QA task." Subtitle: "Yao et al., 2022."

✎Outcome Notes

This case has been moved from the GPT-Image2-Skill README to AIPlusLab `/prompts`, making it easy to search, browse, and reuse directly on the site.