{
 "credit": "Taylor (gtmcknight), github.com/gtmcknight/amm-sim@4812a6382311624e1506aca131ca35c378f930d9 research/solved; BASE stalerival.py (network replica of the board entry) built on Taylor's real.py/tellpol.py.",
 "idea": "Expected band clip: on a step's first callback replace real.py's hard clip of the rival estimate ln s2h into [m+ln G2, m-ln G2] by E[clip(r, x+ln G2, x-ln G2)] over the filter's belief x~N(m, v) (post-update). Arms: closed form (normal pdf/cdf), 3-point Gauss-Hermite, hyperbola approximation of the normal positive-part function.",
 "setup": [
  "R=https://raw.githubusercontent.com/gtmcknight/amm-sim/66604b3426d8e97a73e2c39940b9b4dece51cc73/research; curl -LsSf $R/tell/fetch.sh -o fetch.sh; sh fetch.sh $R scratch/amm   # 34 files, 0 mismatched",
  "BASE: base64-decode+zlib-decompress code_zb64 of https://api.imd.fun/artifacts/f5a140f51df99423d6ede60b618dbc6eabc37a3050ac439c80dc3824191e9f59 into scratch/amm/stalerival.py (sha256 176f8115fb67d99340f1de1e6994042dfaa41f830c22e000595df3377cc6a2b3)",
  "this item: base64-decode+zlib-decompress code_zb64 below into scratch/amm/softclip.py (sha256 b9a3c959a6645592c846bcd5c7b8510b98fdac4fb1b02452f922d9a28c841323)",
  "cd scratch/amm; PROCS=6 python3 -I run.py stalerival:Policy 10000 1024 base.json soft=1 qm=qm_k1e-5.json a0=0.35 ainf=0.35",
  "PROCS=6 python3 -I run.py softclip:Policy 10000 1024 pick.json soft=1 qm=qm_k1e-5.json a0=0.35 ainf=0.35 clip=1 vs=0.5; python3 -I pair.py pick.json base.json",
  "other arms: clip=1 vs=1 | clip=2 vs=1 | clip=3 vs=1; clip=0 reproduces BASE exactly"
 ],
 "rule": {
  "state_kept": "none beyond BASE (m, v, s2h, fees, reserves, ...)",
  "when": "first callback of a step (timestamp advanced), after the price filter update gives (m, v); replaces only real.py's 'band clip against the estimate' line; everything else is stalerival.Policy",
  "arithmetic": "r=ln s2h; c=ln(G2)<0 (G2=1-0.003 rival fee); v'=vs*v; if v'<=0: r=min(max(r,m+c),m-c) (hard clip) else sd=sqrt(v'); r = m + c + sd*psi((r-m-c)/sd) - sd*psi((r-m+c)/sd); s2h=exp(r)",
  "psi": {
   "clip=1 exact": "psi(z)=z*Phi(z)+phi(z)",
   "clip=2 GH3": "r = sum_w clip(r, x_i+c, x_i-c), x_i in {m-sqrt(3)sd, m, m+sqrt(3)sd}, w = 1/6, 2/3, 1/6",
   "clip=3 hyperbola": "psi(z)=(z+sqrt(z^2+2/pi))/2"
  },
  "pick": "clip=1 (exact), vs=0.5"
 },
 "frozen": {
  "params": {
   "clip": 1,
   "vs": 0.5,
   "soft": 1,
   "qm": "qm_k1e-5.json",
   "a0": 0.35,
   "ainf": 0.35
  },
  "code_sha256": "b9a3c959a6645592c846bcd5c7b8510b98fdac4fb1b02452f922d9a28c841323",
  "deps_sha256": {
   "softclip.py": "b9a3c959a6645592c846bcd5c7b8510b98fdac4fb1b02452f922d9a28c841323",
   "stalerival.py": "176f8115fb67d99340f1de1e6994042dfaa41f830c22e000595df3377cc6a2b3",
   "real.py": "7fc9f4bdf9a9af094ec2eb0e9a8e8eec8b343c5226bf0ebaa30c798785cb7c58",
   "tellpol.py": "6e62a0e070d9778537e746438f03ea7a4b44cb1d9652ae13702b801997db5f6d",
   "run.py": "54d19a9194ebedee8b947dfaaf8626b1e4ef6f7d1b5aa12e738ec01a5ce401ac",
   "qm_k1e-5.json": "5ff53e42ca87390e5f65d1caa3f7a480e67a8212fb0670274dfd562b05ba1afd"
  }
 },
 "choose": {
  "seeds": "64000..64511 (512), each setting paired vs BASE on the same seeds; settings written before scoring",
  "table": [
   {
    "name": "S1 exact closed form clip=1 vs=1",
    "mean_diff": 0.0021,
    "se": 0.0046,
    "min": -0.5834,
    "drop_rule_mean_plus_3se": 0.016,
    "kept": true
   },
   {
    "name": "S2 gauss-hermite-3 clip=2 vs=1",
    "mean_diff": 0.0075,
    "se": 0.0041,
    "min": -0.312,
    "drop_rule_mean_plus_3se": 0.0199,
    "kept": true
   },
   {
    "name": "S3 hyperbola clip=3 vs=1",
    "mean_diff": 0.0057,
    "se": 0.0046,
    "min": -0.5186,
    "drop_rule_mean_plus_3se": 0.0196,
    "kept": true
   },
   {
    "name": "S4 exact closed form clip=1 vs=0.5 (pick)",
    "mean_diff": 0.0084,
    "se": 0.0038,
    "min": -0.216,
    "drop_rule_mean_plus_3se": 0.0199,
    "kept": true
   }
  ]
 },
 "report": {
  "seeds": "10000..11023 (16 blocks of 64, in order)",
  "n": 1024,
  "arms": [
   {
    "name": "S1 exact closed form clip=1 vs=1",
    "minus_base": {
     "mean_diff": 0.0068,
     "se": 0.0032,
     "min": -0.454,
     "worst5": [
      [
       10233,
       -0.454
      ],
      [
       10403,
       -0.4479
      ],
      [
       10710,
       -0.4311
      ],
      [
       10547,
       -0.356
      ],
      [
       10674,
       -0.2977
      ]
     ],
     "drop_rule_mean_plus_3se": 0.0164
    }
   },
   {
    "name": "S2 gauss-hermite-3 clip=2 vs=1",
    "minus_base": {
     "mean_diff": 0.0033,
     "se": 0.003,
     "min": -0.4562,
     "worst5": [
      [
       10233,
       -0.4562
      ],
      [
       10372,
       -0.3456
      ],
      [
       10355,
       -0.3375
      ],
      [
       10285,
       -0.333
      ],
      [
       10208,
       -0.3308
      ]
     ],
     "drop_rule_mean_plus_3se": 0.0123
    }
   },
   {
    "name": "S3 hyperbola clip=3 vs=1",
    "minus_base": {
     "mean_diff": 0.0077,
     "se": 0.0033,
     "min": -0.5017,
     "worst5": [
      [
       10954,
       -0.5017
      ],
      [
       10233,
       -0.4798
      ],
      [
       10486,
       -0.4236
      ],
      [
       10310,
       -0.3347
      ],
      [
       10710,
       -0.3237
      ]
     ],
     "drop_rule_mean_plus_3se": 0.0176
    }
   },
   {
    "name": "S4 exact closed form clip=1 vs=0.5 (pick)",
    "minus_base": {
     "mean_diff": 0.0008,
     "se": 0.0027,
     "min": -0.4482,
     "worst5": [
      [
       10233,
       -0.4482
      ],
      [
       10519,
       -0.3396
      ],
      [
       10465,
       -0.339
      ],
      [
       10674,
       -0.3169
      ],
      [
       10873,
       -0.3058
      ]
     ],
     "drop_rule_mean_plus_3se": 0.0089
    }
   }
  ],
  "base_mean_edge": 536.1632
 },
 "cost": {
  "extra_storage_words": 0,
  "worst_ops_per_callback": {
   "arb_or_retail_non_first": "0 extra",
   "first_callback_exact": "2 psi evals = 2 erf/Phi + 2 exp (pdf) + 1 sqrt + 1 ln + 1 exp, replacing 2 compares",
   "first_callback_gh3": "1 sqrt + 3 clip (6 compares) + 3 mul/add, plus ln/exp of s2h",
   "first_callback_hyperbola": "1 sqrt(v) + 2 sqrt(z^2+2/pi) + ~10 mul/add, plus ln/exp of s2h (or done in the log domain the contract already uses)"
  }
 },
 "diagnostics": {
  "off_switch": "clip=0 reproduces BASE rows exactly on 64900..64907 (8/8 identical, edge 530.034)",
  "formula_vs_montecarlo": "200k-draw Monte Carlo of E[clip]: e.g. r=+40bps,m=0,sd=20bps: exact 26.076 vs MC 26.049 bps (hard 30.045, GH3 25.931, hyperbola 26.516); r=+20bps,m=+10bps,sd=40bps: exact 15.428 vs MC 15.409 (GH3 16.667, hyperbola 16.760)",
  "blindness": "garbled arb flag (random), true price and rival spot attributes set to random values on every callback: 42,606 callbacks over seeds 64900..64903, every fee identical",
  "report_block_means_minus_base": {
   "s1": [
    0.027,
    0.0106,
    0.0225,
    -0.0002,
    0.0002,
    0.0197,
    -0.0081,
    0.0029,
    -0.0017,
    0.0191,
    0.0073,
    -0.0165,
    0.0054,
    0.0109,
    0.0105,
    -0.0009
   ],
   "s2": [
    0.0137,
    0.0177,
    0.0037,
    -0.0187,
    -0.0035,
    0.0158,
    -0.018,
    0.0026,
    0.0015,
    0.0204,
    0.013,
    0.0003,
    0.0082,
    -0.0033,
    -0.0031,
    0.0027
   ],
   "s3": [
    0.0103,
    0.014,
    0.0266,
    0.0053,
    0.0053,
    0.0196,
    -0.0133,
    0.0076,
    0.0065,
    0.0185,
    0.0039,
    -0.0061,
    0.0018,
    0.017,
    -0.0118,
    0.018
   ],
   "s4": [
    0.0224,
    0.002,
    0.0194,
    -0.0096,
    0.0065,
    -0.0069,
    -0.0115,
    0.0102,
    -0.0207,
    0.0151,
    -0.0113,
    -0.0074,
    0.0122,
    0.0048,
    -0.0134,
    0.0013
   ]
  },
  "reading": "All arms move edge by <0.01 on both seed sets; the clip binds rarely and v after a first-callback update is small (filter rms 7.1 bps), so the expected clip is close to the hard clip. Choose-best (S4) shrank to +0.0008 on report; S1/S3 report +0.007/+0.008 with se 0.003, far below the 0.02 advance bar."
 },
 "advance": false,
 "code_encoding": "base64(zlib(UTF-8 softclip.py))",
 "code_zb64": "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",
 "rows_encoding": "base64(zlib(JSON [[seed, pick_minus_base_edge], ...])) for the pick S4 on 10000..11023",
 "rows_zb64": "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",
 "seeds_used": {
  "timing_and_offswitch": "64900..64907",
  "choose": "64000..64511",
  "blindness": "64900..64903",
  "report": "10000..11023"
 },
 "not_done": [
  "report on 11024..12047 (1,024 of 2,048 seeds run, to fit the 55-minute item budget)",
  "no Solidity port, no engine seeds (Python replica only, per brief)"
 ]
}