At generation 1000, hardwired embryos approached maximum fitness with positive slope 0.19 vs without-sharing slope 0.05 (30x30 grid)

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• For 20x20 grid, stress-sharing population reached max phenotypic fitness by generation 100; hardwired reached max at generation 1000; without-sharing never reached maxfinding0.858
  Demonstrates benefit of stress sharing across smaller grid complexity.
• Stress sharing populations reached target morphology by generation 400, significantly faster than hardwired (1000) or without-sharing populationsfinding0.796
• Stress-sharing embryo utilized maximum competency of 4725 units until generation 400, then dropped to ~4500; without-sharing used ~100 unitsfinding0.776
  Links competency utilization to stress sharing and morphological success.
• On SWE-bench, harness-benefit peaks at Qwen3-235B (19.3 pp), while weaker Qwen3-32B gains only 4.4 pp and stronger Opus 4.6 gains only 2.6 ppfinding0.745
  Core finding demonstrating non-monotonic relationship between base capability and harness-benefit
• In stress-sharing embryo, average stress reduced to zero at step 40 during sequential target morphogenesis; without-sharing stress settled at non-zero with <55 cells movingfinding0.745
  Shows stress sharing allows perfect formation of part-by-part target patterns.
• Stress-sharing populations reach anatomical targets faster than hardwired or non-sharing populations.finding0.743
  Populations with stress sharing discovered correct morphology by generation 500, vs non-sharing and hardwired (p≪0.01).
• Cells in stress-sharing embryo moved average Euclidean distance ~2500 units until gen 400 then ~2400; without-sharing moved ~200 units (30x30 grid)finding0.742
  Quantifies how stress sharing enables long-range cell movements.
• Five out of 82 external subsystems had χ² values above the maximum null distribution, p=0.00052.finding0.742
  Statistical significance of the prediction after time-flip control.