Эксперимент: Double-Geometric Layerwise ES (#1176)

{ "worked": true, "confidence": 7, "verdict": "Built Double-Geometric Layerwise ES with independent double-geometric integer mutations, analytic Fisher normalization, fading-memory natural-gradient scale adaptation, clipping, and rank utilities. The score/covariance identity matched a finite-difference derivative with absolute error 3.2e-9. On a six-coordinate layerwise-style integer quadratic using 8 fixed seeds and 2160 evaluations per run, the method reached the exact optimum in 100% of runs by iteration 20, versus 25% for Gaussian ES and 87.5% for fixed-radius integer search at the full budget; this is a promising but toy-scale signal. The learned q values reached the lower clip (0.08), so scale adaptation was effective but somewhat collapsed toward sparse mutations.", "metrics": { "baseline": "Gaussian ES: final objective 0.700 +/- 0.415, exact fraction 25%; fixed-radius integer search: 0.100 +/- 0.265, exact fraction 87.5%.", "idea": "Double-geometric ES: final objective 0.000 +/- 0.000, exact fraction 100%; mean objective was 0.0 by iteration 20 across 8 seeds. Score identity absolute error: 3.2e-9." }, "how_to_run": "/home/maxwelhelp/main/bin/python3 run_experiment.py", "files": [ "run_experiment.py", "results.json" ], "limitations": "The mini-experiment is a low-dimensional weighted integer quadratic rather than an actual neural-network quantization or adapter task. Candidate selection is greedy-best rather than a full ES recombination/update, only 8 seeds were used, and the adaptive q values hit the lower bound; no wall-clock, GPU, quantization accuracy, or larger-dimensional scaling study was performed." }