LATIDIA · Investigación

F3-AGI v4.0: Inferencia activa bioenergéticamente regulada, preacondicionamiento metaplásico y síntesis de programas residuales para un razonamiento abstracto eficiente de la muestra

arXiv: 2611.13325v1Tipo de anuncio: nuevo Resumen: el razonamiento abstracto de pocos disparos en ARC-AGI requiere la inducción de transformaciones algorítmicas de 2 a 5 pares de demostración sin entrenamiento previo específico de la tarea. Porque pocos disparos DEM

La noticia

arXiv:2611.13325v1 Announce Type: new Abstract: Few-shot abstract reasoning on ARC-AGI requires inducing algorithmic transformations from 2 to 5 demonstration pairs without task-specific pretraining. Because few-shot demonstrations under-constrain the hypothesis space, program synthesis suffers from the Rashomon effect (predictive multiplicity): heuristic search finds spurious programs with low training error that fail on test inputs, while unconstrained library growth degrades latency. We introduce F3-AGI v4.0, a CPU-native neuro-symbolic and active inference architecture comprising three mechanisms: (1) F3-Metabolic-Optim, combining mistake-gated plasticity, Fisher-EMA preconditioning, and virtual ATP gating; (2) Adaptive Complexity-Guided Program Synthesis (AACR) with a Cycle-3 ResidualPatchSynthesizer and smallest-pair early rejection; and (3) a 512D L2-normalized random projection driving Expected Free Energy (EFE)