Эксперимент: Dual-consistent feature–sample pruning (#1227)

{ "worked": false, "confidence": 9, "verdict": "Built a sparse least-squares FR/L1 prototype with dual-feasibility projection, safe-radius feature screening, dual feature ranking, and separable dual sample scoring. The mathematical checks succeeded: 15/15 safe-screened zero features, no false safe eliminations, KKT error 2.5e-10, and exact column/row mask equivariance. However, the pruning strategy did not provide a reliable validation win: at 50% and 75% reduction its validation MSE was much worse than magnitude-feature plus random-sample pruning, so the claimed practical benefit was not observed.", "metrics": { "baseline": "Validation MSE: 0.1793, 0.1672, 0.1767 at 25%, 50%, 75% reduction; retained features/samples: 45/180, 30/120, 15/60.", "idea": "Validation MSE: 0.1799, 1.6934, 1.3075 at 25%, 50%, 75% reduction; retained features/samples: 45/180, 30/120, 15/60. Safe screening removed 8 features in the full solve." }, "how_to_run": "/home/maxwelhelp/main/bin/python3 run_experiment.py", "files": [ "run_experiment.py" ], "limitations": "This is a synthetic squared-loss convex-head experiment, not CIFAR-10 or a frozen ResNet. The sample-removal score is heuristic and no safe sample-deletion bound or neural-network fine-tuning transfer was tested; only one deterministic dataset/seed was evaluated." }