Criteria for Simple Yet Relevant Models of Computation
yaroslavvb · x · 2026-08-27
The post discusses criteria for simple yet relevant models of computation, suggesting we should ban "reward hacks" found in impractical algorithms. Examples include ignoring constants in Galactic Algorithms, packing unbounded computation into single scalar operations (finite precision only), and assuming unbounded free memory (referencing Bill Dally's model). The author asks for other examples.
More from Research
- GPT-5.6 Builds New Kernel, Achieving 9.7x Speedup on TPU — HuaxiuYaoML · 2026-08-27
- RSI-Exam Benchmark Launches to Test AI Recursive Self-Improvement — HuaxiuYaoML · 2026-08-27
- Gordian Screens 1,327 Targets In Vivo, Accelerating Drug Discovery — juanbenet · 2026-08-27
- Discussion on Multi-Agent Reward Schemes and Convergence — jessi_cata · 2026-08-27
- Why scaling LLMs won't lead to real agency: A 3-tier Embodied AI architecture — Far-Start-1789 · 2026-08-27
- Terence Tao on Human-AI Complementarity: AI Excavates, Humans Recognize — bennash · 2026-08-27