kalomaze: Weak Transfer Persists Because Labs Just Stratify Domains and Pray
kalomaze · x · 2026-09-12
In a thread with StefanGliga, kalomaze argues capability transfer remains weak because the industry isn't explicitly targeting it via better transfer prediction—current practice is overwhelmingly 'stratify across explicit domains and pray.' He adds that humans have bad intuitions about what helps broadly, so labs scale synthetic versions of what SWEs already care about.
More from Research
- Fly connectome trained to play a Flappy Bird–style game — TinfoilTricorn · 2026-09-12
- Falcon Without Floating-Point: PQShield's Fixed-Point Scheme Dodges Side-Channel Leaks — jedisct1 · 2026-09-12
- jd_pressman on alignment: a seed of caring in models is worth tracking — jd_pressman · 2026-09-12
- Intel engineers show wafer-level chiplet testing for co-packaged optics paper — jwt0625 · 2026-09-12
- NVIDIA maps the streaming multi-speaker ASR design space across four architectures, Interspeech 2026 — alexcovo_eth · 2026-09-12
- AI math proofs: verifiable rigor matters, not who ran the AI — abc conjecture as cautionary tale — RexDouglass · 2026-09-12