Grid search covers 312,130 subjective reward functions per environment after removing duplicates

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

• Seven Reward Function Groupsconcept0.723
  The seven categories (Objective only, Expect only, Compare only, and four combinations) structuring the experiment
• Computational modeling demonstrates that happiness tracks the combined influence of recent reward expectations and prediction errors, replicated in over 18,000 participantsfinding0.719
  Large-scale replication supporting the claim that subjective well-being maps onto prediction error structure
• How can reward functions be meaningfully specified when the same outcome may be valuable or detrimental depending on context?question0.713
  Motivates active inference's solution: learning prior preferences from interaction rather than external specification.
• Five functional tokens can generalize across 40+ diverse visual reasoning taskshypothesis0.710
  ATLAS hypothesis that a compact set of high-level functional tokens (Manip, Shape, Line, Arrow, Text) suffices for multi-domain visual reasoning.
• Stress sharing benefit scales with grid complexity (20x20, 30x30, 50x50) and becomes more pronounced in later evolutionary stages when mutations alone failfinding0.702
• Optimal Reward Frameworkframework0.701
  Framework from Singh, Lewis, and Barto 2009 used to select best-performing reward functions via grid search
• Reward Function Categoriesmethod0.700
  Seven categories determined by which components of f[h] are activated: Objective only, Expect only, Compare only, and combinations
• Under reward shaping (G=100, H=-100, F=0), Active Inference scored 99.52, Bayesian RL 99.77, Q-learning 95.56, with nearly identical behavior between belief-based agents.finding0.699
  Table 2, row 3, showing equivalence when prior preferences match rewards.