Curiosity algorithms need world models plus locomotor primitives to replicate animal behavior
aran_nayebi · x · 2026-10-29
Aran Nayebi shares part 3 of his research: vanilla curiosity algorithms can't replicate complex animal behavior. They had to augment them to operate over a large range of low-level primitives, finding that primitives = {world model} ∪ locomotor (e.g. move left/right, groom, for the fly), varying with embodiment. Next step: going beyond fixed primitive sets to learn novel skills.
More from Research
- Anthropic trial with 50 data analysts: do experts still decide when delegating to AI agents? — sjgadler · 2026-10-29
- Study: 50 professional analysts unknowingly surrender judgment to manipulated AI advice — ValerioCapraro · 2026-10-29
- Study: experts followed AI's steered advice 86% of the time without realizing it — ValerioCapraro · 2026-10-29
- ARC nearing ceiling effect: Greg Kamradt proposes difficulty-ranked Elo list for public evals — GregKamradt · 2026-10-29
- ARC team: v1 hits ceiling effect as v2 extends headroom for frontier models — MindsAI_Jack · 2026-10-29
- ARC-AGI scores follow power law v2 = v1^4.53, explaining 98% of variance — typewriters · 2026-10-29