How to Use Stepfun Diligence Check for Output refinement passes

阶跃星辰于2023年4月成立，以“智能阶跃，十倍每个人的可能”为使命。阶跃星辰坚定自研超级模型，积极布局算力、数据等关键资源，发挥算法和人才优势，已完成 Step-1 千亿参数语言大模型和 Step-1V 千亿多模态大模型的研发，在图像理解、多轮指令跟随、数学能力、逻辑推理、文本创作等方面性能达到业界领先水平。 Confirm live details on stepfun.com before production use.

Practical Output refinement passes examples

Example 1

Scenario:
Stepfun Diligence Check — Output refinement passes (pass 1). Context: 阶跃星辰于2023年4月成立，以“智能阶跃，十倍每个人的可能”为使命。阶跃星辰坚定自研超级模型，积极布局算力、数据等关键资源，发挥算法和人才优势，已完成 Step-1 千亿参数语言大模型和 Step-1V 千亿多模态大模型的研发，在图像理解、多轮指令跟随、数学能力、逻辑推理、文本

Objective:
Deliver a reviewable output refinement passes result using Stepfun Diligence Check.

Inputs:
- Verified facts from stepfun.com
- Audience, channel, or technical constraints
- Success criteria and forbidden claims

Workflow:
Open Stepfun Diligence Check → Configure for output refinement passes → Pilot with sample inputs → Review against success criteria → Iterate one axis → Finalize

Requirements:
- Use only verified Stepfun Diligence Check capabilities; do not invent features.
- Confirm live details on stepfun.com before promising volume or pricing.
- Human-review before external publish, send, billing, or compliance use.

Expected output:
A concrete output refinement passes artifact plus a short verification checklist.

Example 2

Scenario:
Stepfun Diligence Check — Output refinement passes (pass 2). Context: 阶跃星辰于2023年4月成立，以“智能阶跃，十倍每个人的可能”为使命。阶跃星辰坚定自研超级模型，积极布局算力、数据等关键资源，发挥算法和人才优势，已完成 Step-1 千亿参数语言大模型和 Step-1V 千亿多模态大模型的研发，在图像理解、多轮指令跟随、数学能力、逻辑推理、文本

Objective:
Deliver a reviewable output refinement passes result using Stepfun Diligence Check.

Inputs:
- Verified facts from stepfun.com
- Audience, channel, or technical constraints
- Success criteria and forbidden claims

Workflow:
Open Stepfun Diligence Check → Configure for output refinement passes → Pilot with sample inputs → Review against success criteria → Iterate one axis → Finalize

Requirements:
- Use only verified Stepfun Diligence Check capabilities; do not invent features.
- Confirm live details on stepfun.com before promising volume or pricing.
- Human-review before external publish, send, billing, or compliance use.

Expected output:
A concrete output refinement passes artifact plus a short verification checklist.

Example 3

Scenario:
Stepfun Diligence Check — Output refinement passes (pass 3). Context: 阶跃星辰于2023年4月成立，以“智能阶跃，十倍每个人的可能”为使命。阶跃星辰坚定自研超级模型，积极布局算力、数据等关键资源，发挥算法和人才优势，已完成 Step-1 千亿参数语言大模型和 Step-1V 千亿多模态大模型的研发，在图像理解、多轮指令跟随、数学能力、逻辑推理、文本

Objective:
Deliver a reviewable output refinement passes result using Stepfun Diligence Check.

Inputs:
- Verified facts from stepfun.com
- Audience, channel, or technical constraints
- Success criteria and forbidden claims

Workflow:
Open Stepfun Diligence Check → Configure for output refinement passes → Pilot with sample inputs → Review against success criteria → Iterate one axis → Finalize

Requirements:
- Use only verified Stepfun Diligence Check capabilities; do not invent features.
- Confirm live details on stepfun.com before promising volume or pricing.
- Human-review before external publish, send, billing, or compliance use.

Expected output:
A concrete output refinement passes artifact plus a short verification checklist.

Example 4

Scenario:
Stepfun Diligence Check — Output refinement passes (pass 4). Context: 阶跃星辰于2023年4月成立，以“智能阶跃，十倍每个人的可能”为使命。阶跃星辰坚定自研超级模型，积极布局算力、数据等关键资源，发挥算法和人才优势，已完成 Step-1 千亿参数语言大模型和 Step-1V 千亿多模态大模型的研发，在图像理解、多轮指令跟随、数学能力、逻辑推理、文本

Objective:
Deliver a reviewable output refinement passes result using Stepfun Diligence Check.

Inputs:
- Verified facts from stepfun.com
- Audience, channel, or technical constraints
- Success criteria and forbidden claims

Workflow:
Open Stepfun Diligence Check → Configure for output refinement passes → Pilot with sample inputs → Review against success criteria → Iterate one axis → Finalize

Requirements:
- Use only verified Stepfun Diligence Check capabilities; do not invent features.
- Confirm live details on stepfun.com before promising volume or pricing.
- Human-review before external publish, send, billing, or compliance use.

Expected output:
A concrete output refinement passes artifact plus a short verification checklist.

Example 5

Scenario:
Stepfun Diligence Check — Output refinement passes (pass 5). Context: 阶跃星辰于2023年4月成立，以“智能阶跃，十倍每个人的可能”为使命。阶跃星辰坚定自研超级模型，积极布局算力、数据等关键资源，发挥算法和人才优势，已完成 Step-1 千亿参数语言大模型和 Step-1V 千亿多模态大模型的研发，在图像理解、多轮指令跟随、数学能力、逻辑推理、文本

Objective:
Deliver a reviewable output refinement passes result using Stepfun Diligence Check.

Inputs:
- Verified facts from stepfun.com
- Audience, channel, or technical constraints
- Success criteria and forbidden claims

Workflow:
Open Stepfun Diligence Check → Configure for output refinement passes → Pilot with sample inputs → Review against success criteria → Iterate one axis → Finalize

Requirements:
- Use only verified Stepfun Diligence Check capabilities; do not invent features.
- Confirm live details on stepfun.com before promising volume or pricing.
- Human-review before external publish, send, billing, or compliance use.

Expected output:
A concrete output refinement passes artifact plus a short verification checklist.

Checklist before you ship

• Confirm the workflow stays inside verified Stepfun Diligence Check capabilities
• Review outputs against stepfun.com when accuracy or pricing claims matter
• Keep a short verification list for any claim you would publish externally