# Эксперимент: Composite Density-Power Loss (#978)

{
  "worked": true,
  "confidence": 9,
  "verdict": "Built the exact categorical composite density-power loss with its normalization term and verified two quantitative predictions. Across alpha=0.1, 0.3, 0.5, 1.0, the observed-score gradient ratio followed (1+alpha)q^alpha with maximum relative error below 6e-7 and log-log slopes matching alpha; the full loss gradient converged toward cross-entropy as alpha approached zero. In the noisy toy classifier, DPD reduced the corrupted/clean per-example gradient ratio from 3.38 to 3.04 and corrupted gradient norm from 4.30 to 2.88, but clean validation accuracy was virtually unchanged (92.17% CE vs 92.00% DPD), so no accuracy win was demonstrated.",
  "metrics": {
    "baseline": "CE: clean validation accuracy 0.9217; corrupted gradient norm 4.3036; corrupted/clean gradient ratio 3.3782",
    "idea": "DPD alpha=0.5: clean validation accuracy 0.9200; corrupted gradient norm 2.8766; corrupted/clean gradient ratio 3.0448; observed-score scaling max relative error <=5.75e-7; alpha-to-zero gradient relative error 0.0030 at alpha=0.01"
  },
  "how_to_run": "/home/maxwelhelp/main/bin/python3 dpd_experiment.py",
  "files": [
    "dpd_experiment.py",
    "results.json"
  ],
  "limitations": "Only a small 2D three-class toy dataset was tested; no CIFAR-100, multi-component architecture, calibration, FLOP/speed measurement, alpha annealing, generalized cross-entropy baseline, or repeated random seeds were evaluated."
}