{
  "key": "same-step",
  "status": "complete",
  "track": "B",
  "measurement_provenance": "All score, error, parity, direction and normalization statistics were measured locally. Scores use pinned run.py and replica2.py; pair.py supplies rounded paired checks and score_stats.py adds full-precision sum/sumsq. Holdout lattice rows are the pinned supplied ref/lattice_qm.json. Gas is an operation-cost estimate only.",
  "idea": "Condition the CURRENT price lattice on the executed input of a known-retail same-step callback (dt=0), using the inverse constant-product router, its Jacobian, an all-own-fill branch, a split branch, and positive-own-fill selection normalization. Marginalize total order size analytically and rival spot with nine conditional residual-quantile nodes. Update before the current quote, leaving threshold 0.9 and all quote tables fixed.",
  "settings": {
    "base": "unchanged thrpol:ThrLattice with thr=0.9, qm=qm_k1e-5.json",
    "final_active_strength": 0.1,
    "threshold": 0.9,
    "lattice_cells": 121,
    "cell_width_bps": 2.5,
    "rival_nodes": 9,
    "order_size_model": {
      "mean_Y": 20.0,
      "log_sd": 1.2,
      "log_mean": 2.2757322735539907,
      "note": "fixed midpoint model, not the hidden per-run size parameter"
    },
    "selection": "Strength 0.1 had the best mean among active fitted settings. Zero update had a higher tuning mean; retain base is the deployment recommendation. The active setting is evaluated to report a negative experiment, not presented as a tuning winner.",
    "settings_count_including_rejected_diagnostic_and_control": 4,
    "prior_training_seeds": [
      62000,
      62015
    ],
    "tuning_seeds": [
      62000,
      62127
    ],
    "confirmation_seeds": [
      62128,
      62255
    ],
    "holdout_seeds": [
      10000,
      12047
    ],
    "freeze": {
      "key": "same-step",
      "strength": 0.1,
      "thr": 0.9,
      "frozen_at_utc": "2026-11-10T09:07:41.906409+00:00",
      "selected_before_holdout": true,
      "selection": "best active conditioning strength on tuning; zero-update baseline had higher tuning mean; active setting retained only to test this idea, not for adoption",
      "files": {
        "same_step.py": "f785f83cd894902e0be01ab3172aa75a4c1bb1065380e82d54763482f39d5ecc",
        "prior_params.py": "c0bf1a39fc2d0e41ae8221a6abae0384b1e51ef9a0cb28a116f9827f01e24cf6",
        "fast_same.py": "be2427438fee93babbc6999d8920494c546e962cee9dae4c6dda4a100a0c83a6",
        "diffuse.c": "aec316062acb6acce223ecd80ce234e03bae3a10cb773df5ad675a042c37fa3a",
        "thrpol.py": "663965707d4f20f81ab0ade8ed510742a330605fd9e889d426db1d920abc8958"
      }
    },
    "prior_fit": {
      "seeds": [
        62000,
        62015
      ],
      "fit": [
        {
          "n": 25003,
          "intercept": -0.34906582803660297,
          "slope": 0.0004370290384403407,
          "sd": 2.697090715777137,
          "residual_nodes": [
            0.3337266901489812,
            0.34280374740996095,
            0.3454125227468292,
            0.3469400380216126,
            0.34798603882755774,
            0.3489997612033124,
            0.3508898990491609,
            0.35420923253597664,
            0.36095001002306526
          ]
        },
        {
          "n": 24980,
          "intercept": 0.36909053718894325,
          "slope": 0.0022350827288267166,
          "sd": 2.7421486734834626,
          "residual_nodes": [
            -0.42890111348641674,
            -0.3962091399485397,
            -0.37853814143663916,
            -0.36859383431636106,
            -0.3636058793138227,
            -0.35853786281638933,
            -0.35070203238387004,
            -0.33670790157439584,
            -0.2909837002755833
          ]
        },
        {
          "n": 14356,
          "intercept": 0.08375662012266054,
          "slope": -0.012504460079855171,
          "sd": 8.151659579050431,
          "residual_nodes": [
            -5.0016405031438795,
            -1.386566269924868,
            -0.7131109041655851,
            -0.30009540725346096,
            -0.0837566201272125,
            0.08935081140746251,
            0.4975467254627557,
            1.1970541961930623,
            5.324794273249519
          ]
        }
      ]
    }
  },
  "likelihood_derivation": {
    "input_units": "isBuy=false: U=amountY, Q=retail size in Y, R1=preY, R2=rival preY. isBuy=true: U=amountX, Q=retail sizeY / hypothesized price, R1=preX, R2=rival preX.",
    "pre_reserves": "preY=reserveY-amountY*(1-askFee) on a buy-X retail order; preX=reserveX-amountX*(1-bidFee) on a sell-X retail order.",
    "router": "G=0.997, g=1-own fee, r=sqrt(g/G), a=r*G/(g+r*G), b=(r*R2-R1)/(g+r*G); U=clip(a*Q+b,0,Q). K=1e6 for both pools.",
    "positive_density": "For u>EPS=1e-4: if u<=max(0,b/(1-a)), f_U(u)=f_Q(u); otherwise f_U(u)=f_Q((u-b)/a)/a. Zero-fill mass is 1-survival_Q(max(EPS,(EPS-b)/a)).",
    "selection_event": "Condition on the already observed trade side AND a positive own callback U>EPS. L(p)=sum_j f_U(u|p,rival_j)/sum_j Pr(U>EPS|p,rival_j), equal prior node weights. The numerator integrates to the denominator. Direction-selection likelihood is deliberately not added.",
    "posterior": "z=sum_i P_i L_i; Pnew_i = P_i (L_i/z)^strength / sum_k P_k (L_k/z)^strength. Recompute current m,v before unchanged table lookup. No smoothing of past callbacks.",
    "rival_prior": "Group by previous predicted retail side, or predicted arb. Regress log(true pre-retail rival/s2h) in bps on price-estimation error in bps on 62000..62015; nine equal-weight residual quantiles provide rival_j conditional on each price cell. Coefficients are frozen in prior_params.py.",
    "limitations": [
      "Conditional rival prior is approximate; residual dependence on price/history beyond a linear mean is ignored.",
      "Nine midpoint quantiles omit extreme residual tails.",
      "The fixed lognormal order law approximates the hidden 19..21 per-run size mean.",
      "Conditioning on callback and direction discards side-selection and silent rival-only-order information."
    ],
    "source_urls": [
      "https://raw.githubusercontent.com/gtmcknight/amm-sim/e65314f88de945f6d055bfc2aab01c482ab6570e/research/solved/replica2.py",
      "https://raw.githubusercontent.com/gtmcknight/amm-sim/e65314f88de945f6d055bfc2aab01c482ab6570e/research/solved/latt.py",
      "https://api.imdnotdead.fun/artifacts/22819a1542e36a609ec3d02a160b4b7aaf345d26d7184ef6517abc35b85cfbfc"
    ]
  },
  "tuning": {
    "zero_update": {
      "paired": {
        "n": 128,
        "mean_diff": 0.0,
        "se": 0.0,
        "sum": 0.0,
        "sumsq": 0.0,
        "worst5": [
          [
            62000,
            0.0
          ],
          [
            62001,
            0.0
          ],
          [
            62002,
            0.0
          ],
          [
            62003,
            0.0
          ],
          [
            62004,
            0.0
          ]
        ]
      },
      "diagnostics": {
        "n": 128,
        "first_callbacks": 895051,
        "misread": 262675,
        "first_trade_misread_rate": 0.29347489696117873,
        "callbacks": 1387408,
        "callback_RMS_log_price_error_bps": 13.292866021566669
      }
    },
    "fitted_strength_0_1": {
      "paired": {
        "n": 128,
        "mean_diff": -0.0010688203053748069,
        "se": 0.0020250249329086677,
        "sum": -0.13680899908797528,
        "sumsq": 0.06680762574921015,
        "worst5": [
          [
            62015,
            -0.13254176567636478
          ],
          [
            62003,
            -0.07782920343174737
          ],
          [
            62008,
            -0.07767178644724027
          ],
          [
            62048,
            -0.0505457250277459
          ],
          [
            62020,
            -0.03927493752473765
          ]
        ]
      },
      "candidate": {
        "n": 128,
        "first_callbacks": 895058,
        "misread": 262688,
        "first_trade_misread_rate": 0.2934871259739592,
        "callbacks": 1387446,
        "callback_RMS_log_price_error_bps": 13.291578808108147
      },
      "base": {
        "n": 128,
        "first_callbacks": 895051,
        "misread": 262675,
        "first_trade_misread_rate": 0.29347489696117873,
        "callbacks": 1387408,
        "callback_RMS_log_price_error_bps": 13.292866021566669
      }
    },
    "fitted_strength_1": {
      "paired": {
        "n": 128,
        "mean_diff": -0.004004201643266203,
        "se": 0.006436580366617203,
        "sum": -0.512537810338074,
        "sumsq": 0.6755313429020205,
        "worst5": [
          [
            62043,
            -0.25800010452627475
          ],
          [
            62072,
            -0.1626913769539442
          ],
          [
            62015,
            -0.14964040421676827
          ],
          [
            62073,
            -0.13681288006654313
          ],
          [
            62020,
            -0.13458034915049666
          ]
        ]
      },
      "candidate": {
        "n": 128,
        "first_callbacks": 895080,
        "misread": 262685,
        "first_trade_misread_rate": 0.29347656075434597,
        "callbacks": 1387486,
        "callback_RMS_log_price_error_bps": 13.29329959125919
      },
      "base": {
        "n": 128,
        "first_callbacks": 895051,
        "misread": 262675,
        "first_trade_misread_rate": 0.29347489696117873,
        "callbacks": 1387408,
        "callback_RMS_log_price_error_bps": 13.292866021566669
      }
    },
    "direction_before_strength_tuning": {
      "seeds": [
        62000,
        62015
      ],
      "trajectory": "unchanged base; proposed posterior on copies, no decision changes; labels in outer observer only",
      "rejected_band_prior_strength_1": {
        "callbacks": 64339,
        "true_retail_callbacks": 64339,
        "before_mse_bps2": 163.42135742746777,
        "after_mse_bps2": 166.20616147933035,
        "delta_mse_bps2": 2.784804051862565,
        "mean_update_dot_truth_error_bps2": -0.3317735841372465,
        "mean_squared_update_bps2": 2.1212568835880723,
        "max_posterior_normalization_error": 2.220446049250313e-16
      },
      "fitted_prior_strength_1": {
        "callbacks": 64339,
        "true_retail_callbacks": 64339,
        "before_mse_bps2": 163.42135742746777,
        "after_mse_bps2": 163.43488483026664,
        "delta_mse_bps2": 0.013527402798868913,
        "mean_update_dot_truth_error_bps2": 0.002188717868158886,
        "mean_squared_update_bps2": 0.017904838535186682,
        "max_posterior_normalization_error": 2.220446049250313e-16
      },
      "interpretation": "Band prior direction was wrong and was rejected. Fitted update has positive alignment with true error but full strength overshoots its small signal. Active strengths 0.1 and 1 were then tested."
    }
  },
  "confirmation": {
    "paired": {
      "n": 128,
      "mean_diff": -0.001545950326493628,
      "se": 0.002029924199165432,
      "sum": -0.1978816417911844,
      "sumsq": 0.0672902628755684,
      "worst5": [
        [
          62173,
          -0.15385515168907204
        ],
        [
          62238,
          -0.07772865082426961
        ],
        [
          62177,
          -0.06534009288668585
        ],
        [
          62187,
          -0.05061395836946758
        ],
        [
          62224,
          -0.04614215013646117
        ]
      ]
    },
    "candidate": {
      "n": 128,
      "first_callbacks": 907717,
      "misread": 260073,
      "first_trade_misread_rate": 0.286513307561718,
      "callbacks": 1424881,
      "callback_RMS_log_price_error_bps": 13.252803822328291
    },
    "base": {
      "n": 128,
      "first_callbacks": 907698,
      "misread": 260070,
      "first_trade_misread_rate": 0.28651599981491643,
      "callbacks": 1424868,
      "callback_RMS_log_price_error_bps": 13.253690224251036
    }
  },
  "checks": {
    "zero_change_equivalence": {
      "seeds": [
        62000,
        62127
      ],
      "n": 128,
      "all_rows_exact": true,
      "callbacks": 1387408
    },
    "helper_parity": [
      {
        "fast": "Base",
        "python": "ThrLattice",
        "seeds": [
          62000,
          62003
        ],
        "callbacks": 41161,
        "exact": true
      },
      {
        "fast": "Candidate",
        "python": "SameStep",
        "seeds": [
          62000,
          62003
        ],
        "callbacks": 41152,
        "exact": true
      }
    ],
    "likelihood_normalization": [
      {
        "side": false,
        "rival": 99.0,
        "numeric_mass": 0.08562882900635646,
        "analytic_mass": 0.08562882900634781,
        "conditional_integral": 1.000000000000101
      },
      {
        "side": false,
        "rival": 100.0,
        "numeric_mass": 1.0,
        "analytic_mass": 1.0,
        "conditional_integral": 1.0
      },
      {
        "side": false,
        "rival": 101.0,
        "numeric_mass": 1.000028134146546,
        "analytic_mass": 1.0,
        "conditional_integral": 1.000028134146546
      },
      {
        "side": true,
        "rival": 99.0,
        "numeric_mass": 1.0000140455444368,
        "analytic_mass": 0.9999999950980032,
        "conditional_integral": 1.0000140504465025
      },
      {
        "side": true,
        "rival": 100.0,
        "numeric_mass": 0.9999998741707562,
        "analytic_mass": 0.9999998741707292,
        "conditional_integral": 1.000000000000027
      },
      {
        "side": true,
        "rival": 101.0,
        "numeric_mass": 0.08688449361720757,
        "analytic_mass": 0.08688449361661599,
        "conditional_integral": 1.0000000000068088
      }
    ],
    "router_inverse_and_jacobian": {
      "draws": 20000,
      "positive_own_fill": 9926,
      "all_own": 9514,
      "split": 412,
      "zero_or_below_callback_threshold": 10074,
      "max_relative_order_reconstruction_error": 1.5211936800457771e-12,
      "max_relative_density_jacobian_error": 3.0155975303440215e-13
    },
    "blindness": {
      "passed": true,
      "scoring_harness": "original run.py Blind",
      "visible_callback": [
        "isBuy",
        "amountX",
        "amountY",
        "timestamp",
        "reserveX",
        "reserveY"
      ],
      "hidden_parameters_passed_to_policy": false,
      "flag_or_true_price_in_policy_state": false,
      "future_callback_used": false,
      "offline_labels": "fit_prior.py and direction_check.py outer observers only; never scoring policy inputs or state",
      "source_url": "https://raw.githubusercontent.com/gtmcknight/amm-sim/e65314f88de945f6d055bfc2aab01c482ab6570e/research/solved/run.py"
    },
    "fixed_tables_sha256": {
      "pol_mid.json": "efbb5ee102db4d312f64565d9bf7232b6becbe02b624164f744a402bbbf2983b",
      "qm_k1e-5.json": "5ff53e42ca87390e5f65d1caa3f7a480e67a8212fb0670274dfd562b05ba1afd"
    },
    "baseline_code_tar_sha256": "11ced75444a1fbf5a5bcead5bc957c7ba7a97d813ba8b1d9f2318c5cc94ff223",
    "baseline_published_code_sha256": "f6504da3b5013be52bd556059e8744e8799f715e2334dbc9e772e057ce8c9287",
    "baseline_archive_integrity_note": "The supplied artifact published code_sha256 does not equal SHA256 of its decoded uncompressed tar. The measured hash is reported separately. thrpol.py is byte-identical to the extracted source and remains unchanged; no fast_thr batch helper was used.",
    "baseline_thrpol_sha256": "663965707d4f20f81ab0ade8ed510742a330605fd9e889d426db1d920abc8958",
    "policy_files_unchanged_since_freeze": true,
    "holdout_passes_per_arm": 1,
    "engine_runs": 0,
    "gates_py": "Not run: this item uses unchanged reference replica/filter/table code plus a callback-parity-checked C loop helper. No holdout seeds consumed for parity."
  },
  "gas_estimate": {
    "measured": false,
    "contract_implemented": false,
    "direct_port_within_245000": false,
    "worst_case_added_same_step_likelihood_operations": {
      "price_cells": 121,
      "rival_nodes_per_cell": 9,
      "router_evaluations": 1089,
      "exp": 2299,
      "log": 2299,
      "sqrt": 2178,
      "erfc": 1089,
      "pow": 121,
      "other_scalar_operations_approx": 30000
    },
    "state": {
      "base_probability_scalars": 121,
      "base_other_state_scalars_approx": 14,
      "additional_persistent_history_bits": 2,
      "additional_production_probability_slots": 0,
      "direct_one_scalar_per_slot_base_slots_approx": 135,
      "normalization_and_diagnostic_caches_need_not_be_persistent": true
    },
    "assumed_gas_per_operation": {
      "exp": 500,
      "log": 500,
      "sqrt": 300,
      "erfc": 1000,
      "pow": 1000,
      "other_scalar": 5
    },
    "added_math_gas_approx": 4312400,
    "note": "Hypothetical operation-cost model, not EVM measurement. No Solidity implementation or fixed-point approximation exists here. Reuse of g-dependent factors and grid exponentials can reduce counts. Base lattice storage, table lookups, convolution, and storage access are excluded; direct port is far above the supplied gas/storage targets."
  },
  "reproduction": {
    "instructions": "REPRODUCTION.md inside code archive; extract into pinned solved folder and use python3 -I reproduce.py --checks, --tune, --holdout. Reference tables/rows obtained from pinned fetch.sh, intentionally excluded from bundle. No installed Python packages; optional C helper with stdlib fallback. Every scoring command capped at 470 seconds; holdout in 64-seed chunks.",
    "source_url": "https://raw.githubusercontent.com/gtmcknight/amm-sim/e65314f88de945f6d055bfc2aab01c482ab6570e/research/solved/README.md",
    "code_files": [
      "same_step.py",
      "prior_params.py",
      "fast_same.py",
      "diffuse.c",
      "thrpol.py",
      "fit_prior.py",
      "direction_check.py",
      "quickcheck.py",
      "parity_fast.py",
      "zero_all.py",
      "likelihood_check.py",
      "router_check.py",
      "score_stats.py",
      "aggregate.py",
      "reproduce.py",
      "freeze_settings.py",
      "REPRODUCTION.md",
      "run.py",
      "pair.py",
      "replica2.py",
      "latt.py",
      "real.py",
      "evalpol.py",
      "par.py"
    ],
    "tar_bytes": 92160
  },
  "code_zb64": "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",
  "code_sha256": "abd406cebb199132632525a11558ab3e71dd6a01990c73c870ea690c2185102b",
  "recommendation": "Retain unchanged thr09 base; tuning and confirmation showed no gain. Do not promote this conditioning to Track A.",
  "holdout_vs_base": {
    "n": 2048,
    "mean_diff": -0.0008372698198050865,
    "se": 0.0007187928975686033,
    "sum": -1.7147285909608172,
    "sumsq": 2.167420214747673,
    "worst5": [
      [
        11237,
        -0.3072184599884622
      ],
      [
        11526,
        -0.28397589889561914
      ],
      [
        10945,
        -0.2698397290336061
      ],
      [
        10058,
        -0.26367830211080445
      ],
      [
        10649,
        -0.23458667052909732
      ]
    ]
  },
  "holdout_vs_lattice_qm": {
    "n": 2048,
    "mean_diff": 0.3220546499906861,
    "se": 0.01067830091469974,
    "sum": 659.5679231809252,
    "sumsq": 690.443562230471,
    "worst5": [
      [
        11487,
        -1.88599765519416
      ],
      [
        10942,
        -1.6601541184661528
      ],
      [
        10982,
        -1.4209691758672989
      ],
      [
        11380,
        -1.238180064450944
      ],
      [
        10990,
        -1.1311688823079749
      ]
    ]
  },
  "base_vs_lattice_qm": {
    "n": 2048,
    "mean_diff": 0.3228919198104912,
    "se": 0.01069159987528695,
    "sum": 661.282651771886,
    "sumsq": 692.7408990952625,
    "worst5": [
      [
        11487,
        -1.9115248303883163
      ],
      [
        10942,
        -1.6076924222966227
      ],
      [
        10982,
        -1.4457581244053017
      ],
      [
        11380,
        -1.2365145499310302
      ],
      [
        10990,
        -1.1504964526710637
      ]
    ]
  },
  "diagnostics": {
    "candidate": {
      "n": 2048,
      "first_callbacks": 14472178,
      "misread": 4179987,
      "first_trade_misread_rate": 0.28882915895589456,
      "callbacks": 22636643,
      "callback_RMS_log_price_error_bps": 13.27693783121005
    },
    "base": {
      "n": 2048,
      "first_callbacks": 14472181,
      "misread": 4180058,
      "first_trade_misread_rate": 0.28883400504733875,
      "callbacks": 22636622,
      "callback_RMS_log_price_error_bps": 13.277152756808208
    },
    "lattice_qm": {
      "n": 2048,
      "first_callbacks": 14315685,
      "misread": 2322695,
      "first_trade_misread_rate": 0.162248261260289,
      "callbacks": 22263118,
      "callback_RMS_log_price_error_bps": 13.64283973652866
    }
  }
}