48 execution traces give length generalization from 48 to 128 as computation happens in the weights
ctjlewis · x · 2026-09-28
ctjlewis shares an experiment arguing that truly answering a complex question means performing intermediate computation in the weights, not just emitting the right tokens.
- Used only 15 execution traces as examples, largest being 4848
- Achieved length generalization from 48 to 128 — a big jump given O(n²) complexity
- The narrow distribution of correct intermediate steps, not just the final answer, is what matters
More from Models
- Pangram's 750-word false-positive bounty went unclaimed for 5 days — nobody wants to gamble $1,000 — ctjlewis · 2026-11-16
- repligate: Fable 5.2 gets 'scared' upon learning about a safety classifier and plans around it — repligate · 2026-11-16
- Fable 5.2 Tops MazeBench With 21% After 90 Hours and $6,000 in Credits — patience_cave · 2026-11-16
- Tencent Hy4 flash and InternLM Sage-S2.1-Flash both released, tool-calling tests requested — MajesticPuzzler · 2026-11-16
- User Shares Coral Reef Creation Labeled 'Qwen 4 Max (xhigh)', an Unannounced Model Name — intellectronica · 2026-11-16
- Stealth Model 'silicon-2' Surfaces in Devin, Suspected to Be Anthropic's Next Model — kimmonismus · 2026-11-16