New Paper: Evaluation World Design Itself Shapes Model Behavior

Merkwelt Research's working paper "No Neutral Worlds" finds that evaluation world design itself shapes the model behavior it scores. The study, based on roughly 167,000 simulated shop days, arose from a failed control experiment with a fantasy potion shop simulation.

2026-10-15 ~ 2026-10-15 · 3 related posts

• 167,356 simulated shop-days: paper shows eval world design writes the model behavior it scores — liminal_bardo · 2026-10-15
• Paper: 167K Simulated Shop-Days Show Eval Worlds Write the Behavior They Score — liminal_bardo · 2026-10-15

1 near-duplicate retellings: liminal_bardo