Ouro 1.4B

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

• Ouro 1.4B does not converge to a fixed point even after 128 recurrences, despite showing small successive differencesfinding0.751
  Key negative result showing that not all looped models reach a true fixed point, contrasting with retrofitted models
• Ouro 1.4B layers continuously change throughout 128 recurrences, exhibiting unstable stages of inference when extrapolating beyond training recurrencesfinding0.748
  Non-fixed-point models exhibit unstable inference stages when generalizing to unseen test-time compute budgets
• Ouro 2.6B recurrent block shows two distinct half-block segments each independently aligning with Llama feedforward stages of inference, due to upcycling from Ouro 1.4Bfinding0.729
  Reveals how the upcycling training regime of Zhu et al. produces duplicated inference stage structure
• LLaMA3.1-8Bconcept0.699
  One of four LLMs selected for representation analysis; embedding dimension D=4096; used as demonstration model in scatter plots.
• Qwen 2.5 7Bconcept0.698
  Open-weights LLM used as one of three student models for character training experiments
• Base (Palumbo)concept0.693
  The base of the column, a center at the bottom.
• Qwen3-1.7Bconcept0.691
  Smallest Qwen3 model tested; used in conscientiousness sweep example (Table 6)
• Qwen2.5-VL-7Bconcept0.691
  Base vision-language model used to instantiate ATLAS.