{ "bench_version": 1, "track": "sequence", "model": "transformer_tiny", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "grid": [ { "config": { "lr": 0.0015, "epochs": 10, "keep": 16 }, "mean": 0.718140721321106, "per_seed": [ 1.2147727012634277, 0.9435746669769287, 0.3435657322406769, 0.3706497848033905 ] }, { "config": { "lr": 0.003, "epochs": 10, "keep": 16 }, "mean": 0.759648285806179, "per_seed": [ 1.2128480672836304, 1.0560415983200073, 0.36093801259994507, 0.4087654650211334 ] }, { "config": { "lr": 0.006, "epochs": 10, "keep": 16 }, "mean": 0.8399795442819595, "per_seed": [ 1.2241541147232056, 0.9892252683639526, 0.5402792096138, 0.6062595844268799 ] } ], "best": { "lr": 0.0015, "epochs": 10, "keep": 16 }, "full": { "per_seed": [ 1.2147727012634277, 0.9435746669769287, 0.3435657322406769, 0.3706497848033905, 0.942135751247406, 0.7555773258209229, 1.2186181545257568, 0.6535401940345764 ] } }, "idea": { "config": { "lr": 0.0015, "epochs": 10, "eta": 2, "keep": 16 }, "per_seed": [ 0.9471153020858765, 0.7885468602180481, 0.3400212824344635, 0.34854480624198914, 0.5234649777412415, 0.3543352484703064, 0.9228858947753906, 0.36787140369415283 ], "details": [ { "seed": 0, "metric": 0.9471153020858765, "keep": 16, "peak_states": 64, "graph_edges": 61, "retained_fraction": 0.5, "history_last": 0.8118170817693074 }, { "seed": 1, "metric": 0.7885468602180481, "keep": 16, "peak_states": 64, "graph_edges": 61, "retained_fraction": 0.5, "history_last": 0.8393829234441121 }, { "seed": 2, "metric": 0.3400212824344635, "keep": 16, "peak_states": 64, "graph_edges": 61, "retained_fraction": 0.5, "history_last": 0.370681676864624 }, { "seed": 3, "metric": 0.34854480624198914, "keep": 16, "peak_states": 64, "graph_edges": 61, "retained_fraction": 0.5, "history_last": 0.3139439268906911 }, { "seed": 4, "metric": 0.5234649777412415, "keep": 16, "peak_states": 64, "graph_edges": 61, "retained_fraction": 0.5, "history_last": 0.48530861457188923 }, { "seed": 5, "metric": 0.3543352484703064, "keep": 16, "peak_states": 64, "graph_edges": 61, "retained_fraction": 0.5, "history_last": 0.335369356473287 }, { "seed": 6, "metric": 0.9228858947753906, "keep": 16, "peak_states": 64, "graph_edges": 61, "retained_fraction": 0.5, "history_last": 0.8587548828125 }, { "seed": 7, "metric": 0.36787140369415283, "keep": 16, "peak_states": 64, "graph_edges": 61, "retained_fraction": 0.5, "history_last": 0.3451614006360372 } ] }, "comparison": { "delta_mean": -0.23120606690645218, "idea_wins": 8, "n_pairs": 8, "per_seed_diffs": [ -0.26765739917755127, -0.15502780675888062, -0.003544449806213379, -0.022104978561401367, -0.41867077350616455, -0.40124207735061646, -0.2957322597503662, -0.2856687903404236 ], "p_value": 0.0081, "mde": 0.13177997997997729, "mde_rel_pct": 16.363998280184262, "verdict": "idea better (significant)", "system_worked": true }, "mechanism_signature": { "prediction": "omitted-token interaction mass predicts output sensitivity", "predicted_mean": 0.09890676662325859, "observed_mean": 0.06862932164222002, "correlation": -0.32015715855609533, "confirmed": false }, "idea_sweep": [ { "config": { "lr": 0.0015, "epochs": 10, "eta": 2, "keep": 16 }, "mean": 0.5740982219576836, "per_seed": [ 0.9471153020858765, 0.7885468602180481, 0.3400212824344635, 0.34854480624198914, 0.5234649777412415, 0.3543352484703064, 0.9228858947753906, 0.36787140369415283 ] }, { "config": { "lr": 0.003, "epochs": 10, "eta": 2, "keep": 16 }, "mean": 0.5850291326642036, "per_seed": [ 0.970460057258606, 0.8036816716194153, 0.3443853259086609, 0.37667813897132874, 0.5293618440628052, 0.35860389471054077, 0.9097486138343811, 0.38731351494789124 ] }, { "config": { "lr": 0.006, "epochs": 10, "eta": 2, "keep": 16 }, "mean": 0.6615562364459038, "per_seed": [ 1.0524933338165283, 1.077573299407959, 0.3749794065952301, 0.4042664170265198, 0.5629213452339172, 0.3612123727798462, 1.0209476947784424, 0.43805602192878723 ] } ], "runtime_sec": 65.80087637901306, "structural_match": "multi-token sequence forecasting with token interaction graph", "state_counts": [ 64, 64, 64, 64, 64, 64, 64, 64 ] }