Learned simulations can be partially observed and lazily-rendered, and still work.

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

• Computer simulations cannot yet provide the level of subtlety or depth to judge the visceral effect of complex building spaces; therefore they cannot replace real-world step-by-step feedback.claim0.800
  Limitation of current simulation technology in capturing life-quality of spaces.
• The strict version of the simulation objective is optimized by the actual time evolution rule that created the training samples.claim0.795
  Equivalence of optimal predictor to the physics of the data.
• Today's Large Language Models have become so good at playing Turing's game that it often takes experts to demonstrate the present limits of their ability to simulate human-like intelligence.claim0.790
  Paper's assessment of current LLM capabilities relative to Turing Test
• Simulations of a generative process (done in the laboratory) are essential to test that the defined process generates coherent complex variable morphology before applying it to real landclaim0.785
  Alexander's methodological justification for using the Guasare simulation studies.
• Learned features reflect the functionality of the model and not just the data distribution, as evidenced by interpretable downstream effects not used in dictionary learningclaim0.773
  Authors argue features are model properties because logit effects and ablations are consistent with feature interpretations
• We evaluate the relative life of a process by making predictions based on thought-experiments or simulations.claim0.770
• It's tricky, because for a typical language model the entity is sort of tricameral: the base simulator, the simulated simulator, and the simulated awareness.quote0.769
  Antra's earlier definitive statement of the tricameral model.
• The upper bound of what can be learned from a dataset is not the most capable trajectory, but the conditional structure of the universe implicated by their sum.claim0.767
  Key insight about predictive learning's potential.