Viewing profile — arpperzhao
arpperzhao
HN member- Joined
- Fri, Aug 21, 2026, 11:39 AM UTC
- HN karma
- 1
- Public activity
- 1 items
- HN profile
- View on Hacker News ↗
About arpperzhao
c programmer
Recent public activity
-
comment
Comment #49597721
Conjecture: Does GLM only handle the prefill stage, compress its output hidden vectors (trained?), and then send them to Qwen for decoding?