{"credit":"Taylor (gtmcknight), github.com/gtmcknight/amm-sim research/solved at 4812a6382311624e1506aca131ca35c378f930d9; replica/tell snapshot 66604b3426d8e97a73e2c39940b9b4dece51cc73. LEANBASE and lean tell from the credited efb8e653f8fd061cecf3e92b186470c9c96c5188523364b60ed101e9672869aa artifact.","parameter_ranges":{"source":"https://raw.githubusercontent.com/gtmcknight/amm-sim/66604b3426d8e97a73e2c39940b9b4dece51cc73/research/solved/replica2.py","draw":"Independent uniform draws from random.Random(10000000+seed), in order size_mean, arrival_rate, sigma. Default 10000 steps; p starts at 100; log-price increment is sigma*gauss(0,1)-sigma^2/2. Retail count is Poisson(rate); order size is lognormal with log-SD 1.2 and arithmetic mean size_mean.","sigma":{"min":0.000882,"max":0.001008,"units":"log-price SD per step"},"rate":{"min":0.6,"max":1.0,"units":"total retail arrivals per step, including orders not filled by our pool"},"mean_size":{"min":19.0,"max":21.0,"units":"Y notional per total retail order"}},"levels":{"method":"Three equal-width bins per parameter. Oracle receives actual run parameters only at initialization and maps each to its bin midpoint. Existing Taylor tables retained; no DP re-solve. This measures a specified approximation, not the fully optimized parameter-information ceiling.","bin_rule":"k=min(2,max(0,int(3*(value-low)/(high-low)))); center=low+(k+0.5)*(high-low)/3; middle level is LEANBASE setting.","sigma":{"boundaries":[0.000924,0.000966],"centers":[0.000903,0.000945,0.000987]},"rate":{"boundaries":[0.7333333333333333,0.8666666666666667],"centers":[0.6666666666666666,0.8,0.9333333333333333]},"mean_size":{"boundaries":[19.666666666666668,20.333333333333332],"centers":[19.333333333333332,20.0,20.666666666666668]},"frozen_arms":{"all":{"matchrho":1,"matchsigma":1,"matchsize":1},"base":{},"rate":{"matchrho":1},"sigma":{"matchsigma":1}},"sigma_action":"Set real.SIG2=sigma_center^2 for each callback, including each silent-step variance increment and arb/retail posterior. Baseline SIG2=0.000945^2.","rate_action":"Set real.RHO=0.26*(1-exp(-rate_center))/(1-exp(-0.8)) for each callback. Taylor tell likelihood multiplies this by the observed grid-hit chance as usual. This scales the prior retail incidence approximately; it does not re-solve phase-specific fees.","size_action":"Set real.LD_M=log(11.12e-4)+log(size_center/20) for each callback. LD_S=1.176 remains fixed. This scales the fitted own-fill displacement law approximately; it does not assume own fills are full retail orders.","table_action":"Same pol_mid.json and qm_k1e-5.json for all levels. Training screened symmetric log-margin shifts delta=sum_i coefficient_i*(level_i-1), applied as max(0,1-(1-fee)*exp(-delta*1e-4)) to both lookup fees, leaving initial _fees unchanged. Each coefficient chosen from -0.5,0.5,1.0 bps in broad screen and -0.1,0,0.1 bps in fine screen. All final coefficients are zero.","selection":"Broad 128 seeds: score.py stopped all nine large-margin arms (mean+3SE<0). Fine independent 128 seeds selected best paired mean per parameter among -0.1,0,+0.1, ties toward zero. All three winners were zero; composed all and all_filter are identical. A further 256 training seeds checked frozen settings. No held-out tuning."},"oracle":{"sigma":{"mean_diff":0.004938048280880819,"se":0.0027345297692516024,"n":2048},"rate":{"mean_diff":-0.00046026887062536326,"se":0.0006494908102335568,"n":2048},"all":{"mean_diff":0.003773726947442907,"se":0.002766355079241978,"n":2048}},"stopped_at_oracle":true,"live":null,"estimator":null,"advance":false,"cost":{"table_bytes":0,"table_bytes_definition":"Additional fee-table bytes relative to LEANBASE","existing_table_json_bytes":{"qm":7892018,"pol":780061},"storage_slots":null,"gas_estimate":null,"note":"Python replica only; unchanged lookup tables. No live arm or Solidity layout exists, so storage/gas are unmeasured and no contract cost is inferred from Python."},"reference_zb64":null,"seeds_used":{"training_broad":[64000,64127],"training_fine":[64128,64255],"training_frozen_check":[64256,64511],"report":[10000,12047],"report_n":2048,"gate_reference_only":[10000,10002],"engine":[],"pairing":"Each reported arm and LEANBASE evaluated once per report seed, separately versus the replica fixed 30 bps rival. Seed pairing shares latent price path; retail RNG streams follow each run configuration as in unmodified replica2.py."},"not_done":["No parameter-specific dynamic-program re-solves: used explicitly defined variance/likelihood matching and screened small quote adjustments. A negative result here does not bound gains from re-solved tables or a richer parameter-conditioned policy.","LIVE estimation and 500/2000/5000-step level-accuracy measurements skipped because best held-out oracle gain is below +0.02.","No Solidity implementation, storage/gas measurements, engine evaluation, leaderboard prediction, or submission; this item is Python replica only."],"measurement":{"metric":"replica edge per 10000-step run; candidate minus LEANBASE","uncertainty":"sample standard deviation of paired per-seed differences / sqrt(n), without rounding inputs","threshold":0.02,"best_oracle_mean_diff":0.004938048280880819,"baseline_options":{"soft":1,"qm":"qm_k1e-5.json","a0":0.35,"ainf":0.35},"upstream_artifact":"https://api.imd.fun/artifacts/efb8e653f8fd061cecf3e92b186470c9c96c5188523364b60ed101e9672869aa"},"reproduction":{"policy_python":"# Credit: Taylor (gtmcknight), amm-sim research/solved; LEANBASE from credited efb8 artifact.\n\"\"\"Parameter-level diagnostic. No hidden trade information enters this policy.\"\"\"\nimport math\nimport leanbase, real\nMID=(0.000945,0.8,20.0)\nLO=(0.000882,0.6,19.0)\nHI=(0.001008,1.0,21.0)\ndef level(v,i):\n    return min(2,max(0,int((v-LO[i])/(HI[i]-LO[i])*3)))\ndef center(k,i):return LO[i]+(k+0.5)*(HI[i]-LO[i])/3\nclass Policy(leanbase.Policy):\n    def __init__(self,T,opt=None):\n        opt=dict(opt or {})\n        self.ks=[float(opt.pop(k,0)) for k in ('ks','kr','kz')]\n        self.matchsigma=bool(opt.pop('matchsigma',0))\n        self.matchrho=bool(opt.pop('matchrho',0))\n        self.matchsize=bool(opt.pop('matchsize',0))\n        self.levels=[1,1,1]\n        self.shift=0.0\n        super().__init__(T,opt)\n    def set_levels(self,levels):\n        self.levels=list(levels)\n        self.shift=sum(k*(l-1) for k,l in zip(self.ks,levels))\n    def lookup(self,w,w2,phs,sd):\n        fb,fa=super().lookup(w,w2,phs,sd)\n        shift=self.shift*1e-4\n        if shift==0:return fb,fa\n        return max(0.,1.-(1.-fb)*math.exp(-shift)),max(0.,1.-(1.-fa)*math.exp(-shift))\n    def on_trade(self,*args):\n        old=(real.SIG2,real.RHO,real.LD_M)\n        if self.matchsigma:real.SIG2=center(self.levels[0],0)**2\n        if self.matchrho:real.RHO=0.26*(1-math.exp(-center(self.levels[1],1)))/(1-math.exp(-0.8))\n        if self.matchsize:real.LD_M=old[2]+math.log(center(self.levels[2],2)/20.)\n        try:return super().on_trade(*args)\n        finally:real.SIG2,real.RHO,real.LD_M=old\n","runner_python":"# Credit: Taylor (gtmcknight), amm-sim research/solved and LEANBASE.\nimport os,sys,json,time,math,statistics,hashlib\nsys.path.insert(0,os.path.dirname(os.path.abspath(__file__)))\nimport replica2 as R,leanbase,param_policy\nfrom evalpol import load\nfrom par import pmap\nfrom run import Blind\nHERE=os.path.dirname(os.path.abspath(__file__))\nclass Oracle(Blind):\n    def init(self,sigma,lam,size):\n        self.inner.set_levels([param_policy.level(v,i) for i,v in enumerate((sigma,lam,size))])\n        return self.inner.init()\ndef work(task):\n    seed,arms=task\n    T=load(HERE+'/pol_mid.json');o=dict(soft=1,qm=HERE+'/qm_k1e-5.json',a0=.35,ainf=.35)\n    row={'seed':seed,'arms':{}}\n    for name,opts in arms.items():\n        if name=='base':w=Blind(leanbase.Policy(T,o))\n        else:w=Oracle(param_policy.Policy(T,dict(o,**opts)))\n        r=R.run(seed,w)\n        row['arms'][name]=r['edge']\n        row['parameters']=[r['sigma'],r['lam'],r['size']]\n    return row\ndef stats(d):\n    return {'mean_diff':statistics.mean(d),'se':statistics.stdev(d)/math.sqrt(len(d)) if len(d)>1 else None,'n':len(d)}\nif __name__=='__main__':\n    cfg,out,s,n=sys.argv[1:];arms=json.load(open(cfg));t=time.monotonic()\n    from multiprocessing import Pool\n    checkpoint=out+'.jsonl'\n    rows=[]\n    if os.path.exists(checkpoint):\n        for line in open(checkpoint):\n            item=json.loads(line)\n            assert item['arms_config']==arms\n            rows.append(item['row'])\n    done={r['seed'] for r in rows}\n    tasks=[(i,arms) for i in range(int(s),int(s)+int(n)) if i not in done]\n    with Pool(int(os.environ.get('PROCS','4'))) as pool, open(checkpoint,'a') as ck:\n        for row in pool.imap_unordered(work,tasks,chunksize=1):\n            rows.append(row)\n            ck.write(json.dumps({'credit':'Taylor (gtmcknight)','arms_config':arms,'row':row})+'\\n');ck.flush()\n            if len(rows)%16==0:print('completed',len(rows),'of',n,'seconds',round(time.monotonic()-t,1),flush=True)\n    rows.sort(key=lambda r:r['seed'])\n    data={'credit':'Taylor (gtmcknight), amm-sim research/solved and LEANBASE','arms':arms,'rows':rows,'seconds':time.monotonic()-t}\n    json.dump(data,open(out,'w'))\n    print('seconds',data['seconds'],flush=True)\n    if 'base' in arms:\n        for a in arms:\n            if a!='base':print(a,stats([r['arms'][a]-r['arms']['base'] for r in rows]),flush=True)\n","input_sha256":{"leanbase.py":"ba015769f7981d7798acbbc9aff8c0e0c59825bb1541d55bcdef49311d7a04bc","lean_tell.py":"b70054f8c0561779e40c4c8df3f41ab08da05b9aef591816ac6c9034beb40a95","tellref.py":"4867a33106ce2597863b8d7c8fb396288954b98f291714801fae8820f1f47aa3","tellpol.py":"7d5961c45bec56292533e65607245ad119b8733c760a5bf28f76a2e07bc3960b","stalerival.py":"d124a171d35b024daed840e5449d352b2b5e50a68ce98c8a9aaafaf3419d23da","real.py":"6a4a0f0e7f8b43d38e5eaa9d8b2e847a0771aee7f89551b122fe89bfe7969b5f","evalpol.py":"79a96efe8c543aee125c937554688b37ec2e7df44df4f3435265bba6eb012bf3","replica2.py":"54380c8755f4e42e9a0426759e918eacd2c5f915b7f04e5d0501c394c5846b40","run.py":"91133e4ef485626226679d73349b5b4e374d9713ac5d7e9d36a39e1afd75933c","par.py":"bf8a1a2e2a4bf9a23bf4622d726fd815702298ad1f9fa0dd17197230cce63ec1","pol_mid.json":"efbb5ee102db4d312f64565d9bf7232b6becbe02b624164f744a402bbbf2983b","qm_k1e-5.json":"5ff53e42ca87390e5f65d1caa3f7a480e67a8212fb0670274dfd562b05ba1afd"},"instructions":"Decode the artifact reference_zb64 as base64 then zlib then JSON (the input bundle is a filename-to-source JSON object, not tar); materialize source files and verify hashes. Obtain pol_mid.json and qm_k1e-5.json from the pinned research/solved URL and verify hashes. Save policy_python as param_policy.py and runner_python as measure.py beside them. Save frozen_arms as frozen.json. Run each 256-seed report chunk with PROCS=4 or 8 and timeout 460 python3 -B -I measure.py frozen.json report-START.json START 256. All runtime files belong in scratch. No third-party packages are required."},"evidence":{"origin":"Measurements executed locally with the pinned Python replica, paired against LEANBASE; these are not upstream performance claims.","paired_differences":{"encoding":"base64(zlib(binary little-endian float64 triples))","columns":["sigma_minus_base","rate_minus_base","all_minus_base"],"seed_order":"10000..12047 inclusive","data":"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","decoded_sha256":"401b6b4d8a5f961e2abd659f2e9d4b9701fcc87da599125140865671a6adb5f2"},"baseline_mean":536.603390956178,"blocks":[{"seeds":[10000,10255],"oracle":{"sigma":{"mean_diff":0.004080320399051818,"se":0.007040534930700741,"n":256},"rate":{"mean_diff":-0.0004540222572990693,"se":0.001967766696529838,"n":256},"all":{"mean_diff":0.0048179089251076945,"se":0.007119810980045465,"n":256}}},{"seeds":[10256,10511],"oracle":{"sigma":{"mean_diff":0.0014281036312882023,"se":0.008735191331415381,"n":256},"rate":{"mean_diff":-0.0014548478701359446,"se":0.0019122524070054499,"n":256},"all":{"mean_diff":-0.0011875500462188437,"se":0.008828887545582748,"n":256}}},{"seeds":[10512,10767],"oracle":{"sigma":{"mean_diff":0.0026508360649093454,"se":0.007866621561398472,"n":256},"rate":{"mean_diff":-0.0016899030810286764,"se":0.0018765567679666247,"n":256},"all":{"mean_diff":5.681823763037386e-05,"se":0.007891818655299232,"n":256}}},{"seeds":[10768,11023],"oracle":{"sigma":{"mean_diff":-0.006464248935797112,"se":0.009402879904203323,"n":256},"rate":{"mean_diff":0.0006486009551252003,"se":0.001797200635781503,"n":256},"all":{"mean_diff":-0.006575889028839654,"se":0.00972058483795569,"n":256}}},{"seeds":[11024,11279],"oracle":{"sigma":{"mean_diff":0.01677302219161181,"se":0.00648245819049432,"n":256},"rate":{"mean_diff":0.00034770661414129655,"se":0.0016099372295213649,"n":256},"all":{"mean_diff":0.0094732907903754,"se":0.006494376468488717,"n":256}}},{"seeds":[11280,11535],"oracle":{"sigma":{"mean_diff":0.018404523852294963,"se":0.0077158528189265755,"n":256},"rate":{"mean_diff":-0.0005260567101517122,"se":0.001555006391258204,"n":256},"all":{"mean_diff":0.019824365135891586,"se":0.007484671837344804,"n":256}}},{"seeds":[11536,11791],"oracle":{"sigma":{"mean_diff":0.004280194365657852,"se":0.006958867143472503,"n":256},"rate":{"mean_diff":0.00324376227370049,"se":0.001915167473923477,"n":256},"all":{"mean_diff":0.005122191244209207,"se":0.007353528237234297,"n":256}}},{"seeds":[11792,12047],"oracle":{"sigma":{"mean_diff":-0.0016483653219703243,"se":0.007214225975809879,"n":256},"rate":{"mean_diff":-0.0037973908893544905,"se":0.0020043922226145594,"n":256},"all":{"mean_diff":-0.001341319678612507,"se":0.007207572710591733,"n":256}}}],"training":[{"stage":"broad","seeds":[64000,64127],"n":128,"arms":{"identity":{"options":{},"mean_diff":0.0,"se":0.0,"n":128},"s_var":{"options":{"matchsigma":1},"mean_diff":0.0015296107638258505,"se":0.008963300355892762,"n":128},"r_rho":{"options":{"matchrho":1},"mean_diff":-0.0019773330577255166,"se":0.0017164487442004608,"n":128},"z_density":{"options":{"matchsize":1},"mean_diff":-0.003172161597202461,"se":0.00303579990671022,"n":128},"s-0.5":{"options":{"matchsigma":1,"ks":-0.5},"mean_diff":-0.1456106867678022,"se":0.019545693018133976,"n":128},"s0.5":{"options":{"matchsigma":1,"ks":0.5},"mean_diff":-0.11532753213936475,"se":0.02146821646602686,"n":128},"s1.0":{"options":{"matchsigma":1,"ks":1.0},"mean_diff":-0.3095906059702247,"se":0.0354624705888951,"n":128},"r-0.5":{"options":{"matchrho":1,"kr":-0.5},"mean_diff":-0.13320069503861065,"se":0.022611028571918913,"n":128},"r0.5":{"options":{"matchrho":1,"kr":0.5},"mean_diff":-0.10770328345596258,"se":0.02224218758652343,"n":128},"r1.0":{"options":{"matchrho":1,"kr":1.0},"mean_diff":-0.2924758517566288,"se":0.03726539499222126,"n":128},"z-0.5":{"options":{"matchsize":1,"kz":-0.5},"mean_diff":-0.14603377308515464,"se":0.021033212197739865,"n":128},"z0.5":{"options":{"matchsize":1,"kz":0.5},"mean_diff":-0.08592895894005359,"se":0.022916141269236755,"n":128},"z1.0":{"options":{"matchsize":1,"kz":1.0},"mean_diff":-0.24726609530966215,"se":0.036081765925697284,"n":128}}},{"stage":"fine","seeds":[64128,64255],"n":128,"arms":{"s_var":{"options":{"matchsigma":1},"mean_diff":-0.004556063377694652,"se":0.009515778856000563,"n":128},"r_rho":{"options":{"matchrho":1},"mean_diff":0.004120861897863204,"se":0.0041187416421741615,"n":128},"z_density":{"options":{"matchsize":1},"mean_diff":0.0015897317396316168,"se":0.0017608926299441073,"n":128},"all_filter":{"options":{"matchsigma":1,"matchrho":1,"matchsize":1},"mean_diff":-0.004061015282407876,"se":0.009600647332504165,"n":128},"s-0.1":{"options":{"matchsigma":1,"ks":-0.1},"mean_diff":-0.0416852114755355,"se":0.012940125583117244,"n":128},"s0.1":{"options":{"matchsigma":1,"ks":0.1},"mean_diff":-0.009608335216967578,"se":0.011785208760639425,"n":128},"r-0.1":{"options":{"matchrho":1,"kr":-0.1},"mean_diff":-0.015449803635571246,"se":0.01018836871473421,"n":128},"r0.1":{"options":{"matchrho":1,"kr":0.1},"mean_diff":-0.030772342489928217,"se":0.01089611011586179,"n":128},"z-0.1":{"options":{"matchsize":1,"kz":-0.1},"mean_diff":-0.03720267492803986,"se":0.009673702062202425,"n":128},"z0.1":{"options":{"matchsize":1,"kz":0.1},"mean_diff":0.0005591537367415356,"se":0.01202751284092917,"n":128}}},{"stage":"confirmation","seeds":[64256,64511],"n":256,"arms":{"all":{"options":{"matchrho":1,"matchsigma":1,"matchsize":1},"mean_diff":0.014926064556400442,"se":0.007573516779101535,"n":256},"rate":{"options":{"matchrho":1},"mean_diff":-0.0001288805610348831,"se":0.0023830622689944895,"n":256},"sigma":{"options":{"matchsigma":1},"mean_diff":0.014323176755411282,"se":0.007235529551939882,"n":256}}}],"checks":{"credit":"Taylor (gtmcknight)","middle_level_identity":{"s_var":{"middle_level_runs":48,"max_edge_difference":0.0},"r_rho":{"middle_level_runs":47,"max_edge_difference":0.0},"z_density":{"middle_level_runs":47,"max_edge_difference":0.0}},"identity_wrapper_n":128,"identity_wrapper_max_diff":0.0,"taylor_gates":{"source":"https://raw.githubusercontent.com/gtmcknight/amm-sim/66604b3426d8e97a73e2c39940b9b4dece51cc73/research/solved/README.md","result":"All 12 displayed reference values reproduced to 6 decimal places. No contract gates claimed.","rows":[[10000,448.261189,443.455598,443.522608,443.66352],[10001,576.779199,571.463409,571.623551,571.977293],[10002,531.886497,525.168028,525.865859,526.411631]]}}}}