360's Papers Split: Understanding vs. Generation
heyshrutimishra · x · 2026-07-14
Continuing a previous research thread: 360's six papers can be divided into two categories: Understanding and Generation.
- Understanding: FG-CLIP 2 can more precisely locate targets like "the red cup on the left"; AML enables agents to accurately click specific buttons or files; MoSA learns objects by observing video motion without manual annotation.
- Generation: RevealLayer can alter just a garment while preserving the background; RefTON naturally composites any outfit onto a photo; NAMI makes this generation process efficient enough for productization.
The main takeaway from these posts is "precise understanding + precise generation," emphasizing that this research is now ready for real-world product integration.
Related event: 360 AI Research Focuses on Production-Grade Foundation Models(3 posts)→
More from Companies & People
- $5.3B Healthcare AI Leader: Domain Expertise Beats Tech — elizabeth · 2026-07-22
- Reformation says AI helped drive 80% of its DTC revenue from full-price sales — omooretweets · 2026-07-22
- OpenAI’s Codex and ChatGPT Work agents reportedly hit 10 million users — Polymarket · 2026-07-22
- Claude Managed Agents demo shared with Vercel in a new presentation — brada · 2026-07-22
- Open-weights models force AI companies to choose between scaling up or opening up — _xjdr · 2026-07-22
- Barr Yaron says AI engineering in 2026 is defined by agents, cost, and evals — AI Engineer · 2026-07-22