Physical World & Real-World ImpactShanghai Huahong Grace Semiconductor (上海華虹宏力), East China University of Science and Technology
Hua Hong Grace and ECUST present WaferSAGE
AIMade by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made
Presented WaferSAGE — wafer defect VQA with a 4B-parameter VLM. On the authors' LLM-judge measurement, the 4B model scored 6.493, recovering about 90% of Gemini-3-Flash's 7.149 and far exceeding the 106B GLM-4.6V (4.750). However, GSPO reinforcement learning contributed only +0.009 over SFT
Why it matters
The authors include an actual semiconductor manufacturer, and the design explicitly assumes on-premises use with no data leaving the site. It means defect report automation is possible with 1–2 in-house GPUs. There is, however, no case of mass-production deployment
Source — Primary sources first; secondary reports are labelled “Press”
Paperarxiv.orghttps://arxiv.org/html/2604.27629Added in Week 35, 2026 · See that week
On the map — ⑭ Physical World line
- Previous stopAI Index 2026
- This recordWaferSAGE
- Next stopBMW says Figure 02 supported production of 30,000+ X3 cars in about 11 months