Sticker-removal success rate stayed near 20% in the early phase of training

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

• Agent achieves approximately 70% sticker-removal success rate by end of 500k training stepsfinding0.892
  Main behavioral result demonstrating the model's efficacy in the mirror-mark task
• Sticker Removal Success Criterionmethod0.818
  Operational definition: hand stays within 2 cm of sticker for 50 consecutive steps (0.5 seconds)
• In a single illustrative episode (seed 2), mean EFE after sticker removal was 12.00 lower than before removal after 500k training stepsfinding0.812
  Qualitative confirmation of EFE drop in trained model vs. untrained model (Δ = +1.70)
• Untrained model (0 training steps) shows no clear EFE difference before and after sticker removal (Δ = +1.70)finding0.804
  Control showing that the EFE signal is learned, not inherent to the architecture
• Mean EFE before sticker removal across 80 evaluations: 79.33 ± 4.34finding0.780
  Baseline EFE when sticker is present, used for comparison
• The sticker-removal behavior induced by the self-prior corresponds to stimulus-elicited intention rather than endogenous intention, aligning with the developmental view of early intentional agencyclaim0.777
  Connects the model's behavior to Zaadnoordijk and Bayne's taxonomy of intentional agency
• EFE decrease after sticker removal is statistically significant (Wilcoxon p = 6.33×10⁻⁹) across 80 evaluationsfinding0.776
  Confirms that EFE systematically decreases after sticker removal, validating the self-prior as internal criterion
• Passing a sticker-bearing latent through the self-prior removes the sticker in reconstruction, confirming distribution favors sticker-free statefinding0.732
  Shows the self-prior's generative distribution rejects sticker-bearing states