{"item":"loss-a (batch e04-price-loss)","credit":"Replica (replica2.py), Gaussian filter, fee rule, qm tables and the solved policy are by Taylor (gtmcknight), github.com/gtmcknight/amm-sim @4812a6382311624e1506aca131ca35c378f930d9 research/solved. qt1 (QMRef), TC (TwoComp) and COMBO are IdentityMD items built on them. This item adds only the oracle diagnostic PricePol and an attribution harness.","gates_output":"seed   perfect-info      gaussian+plain    gaussian+qm       lattice+qm\n10000  448.261189  443.455598  443.522608  443.663520\n10001  576.779199  571.463409  571.623551  571.977293\n10002  531.886497  525.168028  525.865859  526.411631\n","gate_P":{"pass":true,"checkout":"4812a6382311624e1506aca131ca35c378f930d9 (git rev-parse HEAD), gates.py unmodified, CPython 3.12.3","all_12_match_README_at_6dp":true,"qt1_QMRef_10000_10002":"10000 443.101741 444.195330\n10001 571.454483 572.172682\n10002 525.908441 526.274760\n"},"gate_Q":{"live_policy_produced":false,"note":"PricePol is oracle-only by design (reads p); it is a diagnostic ceiling, not a candidate. Nothing was tuned: PricePol has no free parameter (TOL=1e-9 is a float-equality tolerance); choose seeds served the diagnostic, report seeds were run once per policy for paired scores.","fixed_callback_randomization":{"COMBO":[{"seed":64000,"callbacks":9656,"identical_with_arb_s2_j_init_randomized":true,"identical_with_p_also_randomized":true},{"seed":64001,"callbacks":10862,"identical_with_arb_s2_j_init_randomized":true,"identical_with_p_also_randomized":true}],"PricePol":[{"seed":64000,"callbacks":9498,"identical_with_arb_s2_j_init_randomized":true,"identical_with_p_also_randomized":false},{"seed":64001,"callbacks":10689,"identical_with_arb_s2_j_init_randomized":true,"identical_with_p_also_randomized":false}]}},"instrumentation_check":"instrument check OK: 24 runs, all 12 total fields equal (==) between replica2 and replica2i, and DiagCombo == Combo (seeds 64000..64007); additionally all 256 diagnostic seeds' instrumented totals equal the unmodified-replica screen edges (==).","hashes":{"pricepol.py":"4f19c0f0eac7c3e8abd0bddc510dda39ac70088c219d21f4a688011bfb1b240a","combo.py":"c5b711a0b4fb4783e8879a484bf87a427171f4a11aa5eb5ec3513c24ead1a796","twocomp.py":"9d3c1833321493fb51bbde4e1b4dabc2f8978bd6d774ee2af8dc8aed13f0f0af","qmref.py":"1dfc9b7a3c2237ff7a93dd9f0af54b9bdab16f2ce0a7eb00784cb4432598db92","blob.bin":"3ac7cb8811ecfb6a17f499aa4e9b70dd55f7fb6432b3aec3b904e11a3dece38e","replica2i.py":"ef8ab8626ffc6e4560d144a8b1aa690990252b546532d220531f1c1bd9d6e8e9","diagcombo.py":"a573df1e493673f80b213e9bbe1a342ff13d54c18151cadf0f3b50b7b8b7a5d1","diag.py":"6ba4023a9976f997e2275552412488eb108a6cd1a3572a5d038cd9145a14c67b","screen.py":"f0956bdf9725b1bc0a5d257148e98e84c081e312f70ee3bcd8474d8af95cf4f6","instcheck.py":"3c2a742b6371b9d3ba360b4373d35de9038820e4ba73058c72f6f4c2c258bc93","T:research/solved/replica2.py":"6e37b4e61e1ffa5ffb1e43912ddb8659105989aa986eb48fcd1d7608f8dca685","T:research/solved/calib_out_2048.json":"fdf24c7aa2b604d770336a5783276da8ed4d50bca9dbcc02e98b4da58f608d26","gq.py":"1f31422ecae971ece922995c72aeee26fc6114c448c4c70389429db345162481","mkri.py":"385b88ab849405832894692bcc8dd5d612561b0e2e6119c33d6cb36784a9ff95","agg.py":"1d51049aaa83171bc313cde47aa490b46fe732463283ca54129c2572cbb5fbca"},"inputs_verified":{"qt1_artifact":"8a6e303b662e1a5ea9bb47f31acedd3988ef65b3ddb66d074517aa59ad41ca40","TC_artifact":"d3c849db2ac31e567b57e52b80b86e8c24c842e5b4125b9d75c78e3cb07115b9","COMBO_artifact":"d360bea6c8c982717b7b43a8069a87013169c89ee3e4da25622db4d880b936d4","QMS1_artifact":"17d249ab0c33fc44931d9b105cf3f9e3846a560bd518b67310c2a56defacdecc","note":"all four byte sha256s equal their ids; combo.py (c5b711a0...) and twocomp.py (9d3c1833...) extracted from COMBO/TC Code blocks match the sha256s COMBO publishes; blob.bin 3ac7cb88... and qmref.py 1dfc9b7a... fetched from the qt1 reference path COMBO cites match COMBO's published sha256s"},"frozen_spec":{"PricePol":"class PricePol(Combo): at every callback comp=[(1, ln p, 0)], m=ln p, v=0. On a step's first own trade (dt>=1): cnt=0; arb reading lp = ln(s(1-fb)) on buy, ln(s/(1-fa)) on sell; is_arb = |lp - ln p| < 1e-9 (COMBO's pi rule under an sd-0 belief at ln p collapses to this); s2h clipped to [0.997p, p/0.997]; synced=False. Non-arb trades: router-identity rival update and cnt/phase exactly as COMBO. Quote: QMRef.lookup(w=(ln s - ln p)1e4, w2=(ln s2h - ln p)1e4, phase, sd=0) on the QMS1 qm tables; fees clamped [0,0.1]. Mixture path never taken.","pricepol_sha256":"4f19c0f0eac7c3e8abd0bddc510dda39ac70088c219d21f4a688011bfb1b240a","replica":"replica2.run(seed, policy) defaults (10,000 steps), unmodified"},"choose":{"seeds":"64000..65023","PricePol-COMBO":{"n":1024,"mean_diff":3.687392,"se":0.032611,"worst5":[[64003,0.531271],[64132,0.658904],[64020,0.906488],[64322,0.919341],[65013,0.984646]],"wins":1024,"sum_d":3775.88972,"sum_d2":15037.24825},"PricePol-qt1":{"n":1024,"mean_diff":4.296997,"se":0.035137,"worst5":[[64322,0.450973],[64020,0.973867],[64509,1.011999],[65013,1.197804],[64084,1.740365]],"wins":1024,"sum_d":4400.124878,"sum_d2":20200.615551},"COMBO-qt1":{"n":1024,"mean_diff":0.609605,"se":0.019401,"worst5":[[64101,-2.776983],[64845,-1.163868],[64374,-1.023434],[64795,-0.983171],[64555,-0.946189]],"wins":875,"sum_d":624.235158,"sum_d2":774.814573},"note":"calib_out_2048.json has HA only on 10000..12047, so no HA pair here","mean_edge":{"PricePol":540.017492,"COMBO":536.3301,"qt1":535.720495}},"report":{"seeds":"10000..12047","PricePol-HA":{"n":2048,"mean_diff":8.021774,"se":0.045883,"worst5":[[10481,3.072917],[10905,3.324188],[10618,3.353896],[11902,3.40083],[10458,3.410866]],"wins":2048,"sum_d":16428.59363,"sum_d2":140612.164087},"PricePol-COMBO":{"n":2048,"mean_diff":3.718987,"se":0.024144,"worst5":[[11701,-1.411865],[11519,-0.802234],[10923,0.333407],[11530,0.729521],[10125,0.919702]],"wins":2046,"sum_d":7616.485153,"sum_d2":30769.446975},"PricePol-qt1":{"n":2048,"mean_diff":4.316378,"se":0.025937,"worst5":[[11376,0.050789],[11519,0.103583],[11808,0.686741],[11062,0.740355],[11701,0.947237]],"wins":2048,"sum_d":8839.941846,"sum_d2":40976.834391},"COMBO-HA":{"n":2048,"mean_diff":4.302787,"se":0.037598,"worst5":[[10146,-1.860404],[10618,-0.141834],[11031,0.316455],[10940,0.415619],[11828,0.625383]],"wins":2046,"sum_d":8812.108477,"sum_d2":43842.841918},"COMBO-qt1":{"n":2048,"mean_diff":0.597391,"se":0.013388,"worst5":[[10146,-3.552497],[11062,-2.334849],[11808,-1.822738],[10519,-1.782691],[11376,-1.750296]],"wins":1753,"sum_d":1223.456693,"sum_d2":1482.295158},"qt1-HA":{"n":2048,"mean_diff":3.705396,"se":0.038086,"worst5":[[10618,-1.633785],[11031,-0.966923],[11181,-0.198083],[11275,-0.179574],[10447,-0.014874]],"wins":2042,"sum_d":7588.651784,"sum_d2":34200.087915},"mean_edge":{"PricePol":537.579209,"COMBO":533.860222,"qt1":533.262831,"HA":529.557434}},"diagnostic":{"seeds":"64000..64255","n_seeds":256,"steps":2560000,"units":"replica edge (Y) per step; gap = PricePol minus COMBO at the same step t of the same seed","total_gap":945.343,"total_gap_per_seed":3.692748,"total_gap_split":{"retail_edge":363.957,"arb_edge":581.387,"fees_(subset_of_edge_view)":941.794},"mean_edge":{"COMBO":527.355393,"PricePol":531.048141},"buckets":{"phase":{"hard_arb":{"count":96466,"mean_gap_per_step":-0.00015454,"share":-0.0158,"gap_sum":-14.907,"retail_part":-9.088,"arb_part":-5.819,"fee_part":262.073},"hard_retail":{"count":285597,"mean_gap_per_step":0.00096977,"share":0.293,"gap_sum":276.963,"retail_part":104.274,"arb_part":172.69,"fee_part":112.658},"ambiguous":{"count":2177247,"mean_gap_per_step":0.0003138,"share":0.7227,"gap_sum":683.225,"retail_part":268.823,"arb_part":414.401,"fee_part":567.622},"none":{"count":690,"mean_gap_per_step":9.123e-05,"share":0.0001,"gap_sum":0.063,"retail_part":-0.052,"arb_part":0.115,"fee_part":-0.559}},"silent":{"0":{"count":1793277,"mean_gap_per_step":0.00040561,"share":0.7694,"gap_sum":727.368,"retail_part":102.59,"arb_part":624.778,"fee_part":549.446},"1":{"count":435198,"mean_gap_per_step":0.00024938,"share":0.1148,"gap_sum":108.531,"retail_part":131.471,"arb_part":-22.94,"fee_part":243.195},"2-3":{"count":259426,"mean_gap_per_step":0.00027554,"share":0.0756,"gap_sum":71.482,"retail_part":84.536,"arb_part":-13.053,"fee_part":120.598},"4-9":{"count":69764,"mean_gap_per_step":0.00041407,"share":0.0306,"gap_sum":28.887,"retail_part":35.523,"arb_part":-6.636,"fee_part":25.279},"10+":{"count":2335,"mean_gap_per_step":0.00388638,"share":0.0096,"gap_sum":9.075,"retail_part":9.837,"arb_part":-0.762,"fee_part":3.276}},"size":{"|z|<1":{"count":1358862,"mean_gap_per_step":0.00031808,"share":0.4572,"gap_sum":432.233,"retail_part":128.943,"arb_part":303.29,"fee_part":56.877},"1-2":{"count":339581,"mean_gap_per_step":0.00129106,"share":0.4638,"gap_sum":438.42,"retail_part":59.71,"arb_part":378.71,"fee_part":-521.827},">2":{"count":94628,"mean_gap_per_step":0.00028819,"share":0.0288,"gap_sum":27.27,"retail_part":52.776,"arb_part":-25.505,"fee_part":278.443},"no_own_trade":{"count":766929,"mean_gap_per_step":6.183e-05,"share":0.0502,"gap_sum":47.42,"retail_part":122.528,"arb_part":-75.108,"fee_part":1128.301}},"call":{"right":{"count":1503983,"mean_gap_per_step":0.00045653,"share":0.7263,"gap_sum":686.611,"retail_part":159.771,"arb_part":526.841,"fee_part":-42.65},"wrong":{"count":289088,"mean_gap_per_step":0.00073096,"share":0.2235,"gap_sum":211.312,"retail_part":81.658,"arb_part":129.654,"fee_part":-143.857},"none":{"count":766929,"mean_gap_per_step":6.183e-05,"share":0.0502,"gap_sum":47.42,"retail_part":122.528,"arb_part":-75.108,"fee_part":1128.301}},"err":{"0-2.5":{"count":828876,"mean_gap_per_step":-6.208e-05,"share":-0.0544,"gap_sum":-51.456,"retail_part":-24.609,"arb_part":-26.847,"fee_part":615.327},"2.5-5":{"count":320676,"mean_gap_per_step":5.094e-05,"share":0.0173,"gap_sum":16.336,"retail_part":43.543,"arb_part":-27.207,"fee_part":130.839},"5-10":{"count":463204,"mean_gap_per_step":0.00016791,"share":0.0823,"gap_sum":77.775,"retail_part":76.205,"arb_part":1.57,"fee_part":149.606},"10+":{"count":946788,"mean_gap_per_step":0.00095331,"share":0.9548,"gap_sum":902.579,"retail_part":268.708,"arb_part":633.87,"fee_part":46.062},"none":{"count":456,"mean_gap_per_step":0.00024108,"share":0.0001,"gap_sum":0.11,"retail_part":0.11,"arb_part":0.0,"fee_part":-0.039}}},"joint_phase_x_silent":{"hard_arb|0":{"count":85298,"mean_gap_per_step":-0.00014804,"share":-0.0134},"hard_arb|1":{"count":5410,"mean_gap_per_step":-0.00015465,"share":-0.0009},"hard_arb|2-3":{"count":3848,"mean_gap_per_step":-0.00038242,"share":-0.0016},"hard_arb|4-9":{"count":1713,"mean_gap_per_step":-2.505e-05,"share":-0.0},"hard_arb|10+":{"count":197,"mean_gap_per_step":0.00036334,"share":0.0001},"hard_retail|0":{"count":206824,"mean_gap_per_step":0.00114331,"share":0.2501},"hard_retail|1":{"count":43366,"mean_gap_per_step":0.00039017,"share":0.0179},"hard_retail|2-3":{"count":27004,"mean_gap_per_step":0.00052943,"share":0.0151},"hard_retail|4-9":{"count":8089,"mean_gap_per_step":0.00086167,"share":0.0074},"hard_retail|10+":{"count":314,"mean_gap_per_step":0.00736594,"share":0.0024},"ambiguous|0":{"count":1500754,"mean_gap_per_step":0.00033544,"share":0.5325},"ambiguous|1":{"count":386260,"mean_gap_per_step":0.00023959,"share":0.0979},"ambiguous|2-3":{"count":228477,"mean_gap_per_step":0.0002565,"share":0.062},"ambiguous|4-9":{"count":59932,"mean_gap_per_step":0.00036657,"share":0.0232},"ambiguous|10+":{"count":1824,"mean_gap_per_step":0.00366788,"share":0.0071},"none|0":{"count":401,"mean_gap_per_step":0.00028904,"share":0.0001},"none|1":{"count":162,"mean_gap_per_step":-0.00059501,"share":-0.0001},"none|2-3":{"count":97,"mean_gap_per_step":0.0005433,"share":0.0001},"none|4-9":{"count":30,"mean_gap_per_step":-0.00030875,"share":-0.0}},"wrong_vs_right_call_share":{"wrong":0.2235,"right":0.7263,"none(no own trade in step)":0.0502},"top5_buckets_by_share":[["err:10+",0.9548],["silent:0",0.7694],["call:right",0.7263],["phase:ambiguous",0.7227],["size:1-2",0.4638]],"combo_first_trade_calls":{"n":1793071,"wrong":289088,"wrong_rate":0.1612},"pricepol_first_trade_call_vs_true_flag":{"n":1787791,"agree":1787791,"max_abs_lp_minus_lnp_on_arb":8.881784197001252e-16,"min_abs_lp_minus_lnp_on_retail":2.7240662805461113e-05}},"findings":["Gate P passed: unmodified gates.py at 4812a638 printed all 12 README values at 6 dp; QMRef (qt1) gave 443.101741 / 571.454483 / 525.908441 and COMBO gave 444.195330 / 572.172682 / 526.274760 on 10000..10002, as the brief requires.","PricePol vs HA on the report seeds 10000..12047 (n=2048): +8.021774 (se 0.045883), better on all 2048 seeds. The requester expected about +8.50. That figure is Taylor's README row 'solved policy, rival tracked from own trades' (true flag, plain pol_mid tables). PricePol quotes through COMBO's table instead: the QMS1-compressed qm_k1e-5 tables at sd level 0. The README says level 0 of the kappa=1e-5 tables differs from pol_mid in about a quarter of the cells. The 0.48 difference is consistent with that table change but was not isolated.","PricePol vs COMBO on the report seeds: +3.718987 (se 0.024144), better on 2046/2048 seeds. PricePol vs qt1: +4.316378 (se 0.025937), better on 2048/2048. On the choose seeds 64000..65023: PricePol-COMBO +3.687392 (se 0.032611), PricePol-qt1 +4.296997 (se 0.035137).","COMBO-qt1 on the report seeds is +0.597391 (se 0.013388, 1753 wins), the same as COMBO's published figure. COMBO-HA is +4.302787 (se 0.037598) and qt1-HA is +3.705396 (se 0.038086).","PricePol's sd-0 first-trade call agreed with the true arb flag on all 1,787,791 first trades on 64000..64255. |lp - ln p| was at most 8.9e-16 on arb trades and at least 2.7e-5 on retail trades. So PricePol has both the true flag and the true price; it is not a flag-only oracle.","COMBO's first-trade call is wrong on 16.12% of first trades on 64000..64255 (289,088 of 1,793,071), matching the README's 16%.","Attribution, 64000..64255: the gap totals 945.343, or 3.692748 per seed. Of that, 581.387 is arb-loss edge and 363.957 is retail edge, so 61.5% of the gap is in arbitrage losses.","Steps where COMBO's first-trade call was wrong carry 22.35% of the gap. Steps where it was right carry 72.63%, and steps with no own trade carry 5.02%. Per step, wrong-call steps lose 1.6x more than right-call steps (0.000731 vs 0.000457), but right calls are 5.2x as frequent. Fixing the arb/retail call alone therefore addresses at most about a quarter of the per-step gap measured this way. Most of the gap is belief imprecision in steps where the call was right.","Belief error at the quote in force is the sharpest split. Steps whose last quote was set with |m - ln p| >= 10 bps are 37% of steps and carry 95.48% of the gap: 902.579, of which 633.870 is arb loss. Steps with error under 2.5 bps carry -5.44%, so there COMBO slightly beats PricePol.","By phase at the quote: 'ambiguous' (last first-trade pi in (0.005, 0.995)) is 85% of steps and 72.27% of the gap. 'Hard retail' (pi <= 0.005) is 11% of steps and 29.30% of the gap, with the highest per-step gap (0.000970). 'Hard arb' carries -1.58%.","By silent steps before the step: 0 silent steps carry 76.94% of the gap, nearly all of it arb loss (624.778 of 727.368). With 1 or more silent steps the gap is mostly retail edge. 10+ silent steps have the largest per-step gap (0.00389) but only 0.96% share.","Joint phase x silent: ambiguous|0 holds 53.25% of the gap and hard_retail|0 holds 25.01%. Those two cells cover 78% of the gap.","By first own trade size against the retail size law: |z| between 1 and 2 is 13% of steps but 46.38% of the gap, with a 4x higher per-step gap than |z|<1 (0.001291 vs 0.000318). |z|<1 holds 45.72% and |z|>2 holds 2.88%.","Top 5 buckets by share, across families, so they overlap: err:10+ 0.9548, silent:0 0.7694, call:right 0.7263, phase:ambiguous 0.7227, size:1-2 0.4638."],"limitations":["Replica only (Taylor's replica2, not the Rust engine). No contract, engine run or submission, per the brief. Replica levels do not convert to board scores.","PricePol is an oracle diagnostic. It reads p, so Gate Q's live-policy test does not apply to it; the randomization test shows it reads p and nothing else. No live policy was produced, chosen or tuned, so nothing was selected on any window.","The per-step gap compares two different trajectories at the same step index of the same seed. Price path and order draws are shared, but pool states diverge. Bucket labels come from COMBO's trajectory only. A step's gap is partly caused by earlier steps' quotes (inventory carries over), so the shares are associations, not causal effects.","Bucket definitions are this item's reading of the brief. 'Phase at the quote' and 'belief error' use COMBO's most recent callback before step t: the category of its latest first-trade pi, and |m - ln p| right after that callback, at that callback's price. 'Silent steps' is t minus the step of the last own trade, minus 1. Size and call come from COMBO's first own trade in step t; steps without one are bucketed no_own_trade / none.","The fee_part column per bucket is the gap in fees collected (the replica's fees1, valued at the fair price). It is a separate accounting view, not a third additive component; gap_sum = retail_part + arb_part.","The source of the 0.48 shortfall against the requester's expected +8.50 (table difference: pol_mid vs QMS1 qm level 0) is inferred, not measured. track.py was not re-run here.","Report and choose pairs against qt1 use qt1 recomputed here with QMRef on every seed, not published edges. COMBO-qt1 on the report seeds equals COMBO's published +0.597391."],"source_bundle":{"encoding":"json {path:text}, sort_keys, zlib level 9, base64.b85encode","files":["pricepol.py","diagcombo.py","diag.py","screen.py","instcheck.py","gq.py","mkri.py","agg.py"],"decode":"python3 -I -c \"import json,zlib,base64,sys;r=json.load(open(sys.argv[1]));b=json.loads(zlib.decompress(base64.b85decode(r['source_bundle']['data'])));[open(k,'w').write(v) for k,v in b.items()]\" result.json","data":"c-q}u36tAKcK$0JB^v-1iQzFP$5gF%q>Wv%R@#wr70E(@K$8Ov2|xfihh+Wtyzg~4&;U8I?A@d)(XKf_pRZrPes|;Zvv9rk^T%fs@odij`(eE<<T|XRD5FBkv=q-poP_H%E2|_DNm^y1kmYVu`5($G6>cH(O%jD-`6%9pkDIIz>uMX_rpfxI@_ex?!?m1<{IR;p(y{3OB-D)J&A+`@GtSdwn`cEO%Ez)NFlbL~!|J9dK5dd^Pb|Yyo*wtaO<3Mg!8~0R*;eG#g5HGg`<7nk>Ghj8@5Pi~+#pyb8yN(iPvB;oJL!6U9u_jKF2y^v;>Jh9DJxp7Gf&^W`T5^pz0?)DDc2X#5e*^3BDy)^kz#5+Ph+_fn=Fjotv8vcg8s$S`STa`ydp$w2wlQ`hgsY$;n87~Z5+?@vs|XG>4$%`bV=ROz{jYiyYVhXSe6o)5ouOou(&5&C*I`_p{OTHY7x)M1@)6^q(WA^f|||KDC>zl?}?8?s%0>pDBCVG$Ls0WJa3*qRzt^|r#B&1P|TcC(jpfVL60(COJ~9BDF8wZq04K~SR*_|A?-(y?W$ljI6mj4(n>|L!YH>%N&|}NR5(BW;1G_kegs@eS-C}+uBCf=JQxgm;`C%NI`(K?*KgbP!vWV1F}6pb0-l)h+pN@^7=(qJjaxFp7;t{18K2Xzo(R)eYzXw&13Y-T`q860JW>5T?G8d{sYl@kYV-3jDP-)1M4~<M&L)DH8)94fVo86nPfh=Ki%XG;sRrOuZ0Wh&ZQTcM^3W3jgY6MUQ#D|T5pf{=@?%lCXiTm9V(W?Tz7wPUer`&&Lm<f_iO){zOavV*XDh=rh?CWd-ivIP5&;w5i4<tp>mr);$eJ|uxJg5<_~+bbMOmFtpV^GaS9N?q!-WBgmt4{TB|OF6Y;v+7bU61(T2e2y>=$CtSweRjY8S)}WCxIC;77wzr{#CzP-~(<-~k&EZkKV0ll9K-$mgorHQ*In_Kthg@w&;f5(oA(pA{^}VG{OzknoresPnc+BKbDk^snB${?i+3p$P`W3Zf;dvs>W|w%U9gsx~${93HFP7Gs&V!XZw9Z9H@w!9vB4KM3MMkMtw>O!+>})EVtnMqCEf7*^cKLOSiKnQ)jLOla8#ZhAolL#!Y$%jM!`p$cX$)X;vRT_EuR{@p^P4Oa`1#5!6Sx}md22l5+mjaIG2d>!p-P=#xIQVi8h`jx%dn~z$xMnWSq(r?7FHMxytRoua{D&&c|Oiu~S&(nVtefnpMcSFL*qmvWbo9w;>H){!{u4Hr2-3T@<*B0}XHxs-~9sXLTG=c=9TOv-iKPVFjno|$5ipJ)4n*Z0sObBuR$m<YQR)w2VY<FcP<i}mO5!H<pyL6io%S&Y3H~NZFZx|f3D8sE}Ak=1Ps0eH&H5auYBoL!)R<CEl;)GsFoVFU}xz(sA>Sq!!>oqG1m*PW4`jl20+6!uwh?jISi-_^<QcoU<{yGanh|2?T5kFC(ty7@;ZBN`W6D}e@spPh7X&t;&0^p@z%8F!KNP3-1U1o&#fp7n}@Y=mFcEaHDR-Hih&XR?AJ{5QLcb=wYCG%2Cx4YS3(E<lbpausT!mUqQSW%Vt32AOlPJBvE97wSDtk9_;hQE7=98|0ZNq)~D`>S0j6=1ck+&hENTGB=&MJhURShW$j>o5;;Sp>jJEunZ#N4niPL%^7l8<H)5$x;cSc0-zIw+F4MCyw_dB_^R|wp#bZ7%YH*3zJPiYHejzk4$w~EW34MQ@4_`TX)QLM73|OU%t9#?#;-8F5nQraX<rtb@Q{D8?Pj5ll3m!l@3d)l+Fs4NQycmY~~D5F~rZP&qDOLf5AnA=hW0%E_M2(ru*g7FW+;IVSj`_FGgzcAiGb4st98>3CY6^w=ZCspvirarE3cidvsY83ZDMRKVfkElRiJ3^oIijQQem?vD`(svMR09yb4LNx@G-=(bPGAIgAQE{<B*}lbRs+2$2SY^+1%kXmKYxv#uV+6W`Osbwm2K#=I@R^vrZ>01fPnI>37Pd^=%MrpafmF&3UZ3IJ<Qp#at<k2&lZ<#tgYS2QwE*6co9mZV$vLC6HU7|Q-BX=>Z>!6l)EN>qTuF%n6;#QQw`U~|j-rbg{dli$K(oaM%<)+DyDc;77zMMKbHe!*hOO;d#pU|R=Fbs|1gv>qmMbF#Sti~N-XUbo{@+v;`h?H%rRYkkIPu~n8glGJruTR?nlriam+g=&J`(4)=m>}@yHNMPK`$Ej9>A10#zK=ZS%K2P&}WmqSRxL~yjN%R;l+Q{*SL%OiiK&TN;oYk8JX@J7=8%JP)eo)1Re!{0OFjb@?0H2UKuygo<3U+ByaZe-mA)tl(U!I`75@>ZPXy@QG(3+OmCh4vetX_$*vXldAEGm($+WLa6Bs;8qfE4OI<y~YpTGuHuL0790Z8$HXWo21?N&>%Pl~8kXts>R~<ZWiYUv2aAG9V*oN&F+oR@^kWOXR)R?x2~x4NqG+H7aOhH675GI=kE~Q`pLFYlxOA^;G{BT$-?HApJz6^Jxm0F9Z)kEn_}Zc1M;0!Ce`!RAA~1%%S|OB}6^wObL;Nzb3|8i;vG=JXN6jv9M(k^F%)*fA;#dh56I}SmwB7tC4^cEZ|tdBl7f~nv<o}{*+(Zw3eR)x_7XAtkFQ_b~AO1R|%Q^O1kPJJ<%=(KK6<)8uY26VO<myVbu_)sim!L`X0Ao%Rwo5w%MB(TUQl3sTh-if!LOUAY7A@uM2p7k}v^@mtd0ze9nVSng?0R>dJ)WPm9luy7c$Wo7CPhn(N@SS~Hq}>9l7acV|{6+n;~>`PDlx)fy#!{s>M8UddZyz3WG1>FAe)SYvb}qg(yH{-__ew?(2Km4_Cl%91u{!)}h%M7Y{xx`InJjDt!(RIZje0Fz-OC}kQ^59PM?m**$+p2gCIH`gZ+p$n8Q+m!Kco0qPdq{jj-AX!vS-_wqIzV1f{>caE<SgJR8Da#wliq^(ww&;4axjs027)jUh{i6du+as*({ba9{43}mYZKc@e?W)K=5%%ab5aV%LOIO`CNh(Q>8l|ts#6&1SZPRL1yRr3vZWpP`#|oo){*8QyleHvc7)DH?7EILVKyE{#ZL+fLhr(Tkw4p){j{8g!eUy7Fl^ojmn5Bkut;L^l*N<p}39<hY?#eO=$qY_5l`MLKWP>Pn*o}`{QH9G5%?ksUMIK{J5X~jgqafyB`TSkW>E9D;(pq=C*+)(JQMldq%VaD5cziw_g{R~5(ResK9UaTz$>21M!r^!rh2xWGe0ILN7!TqL(b+6te5{7T{p;&%xuRL#UkUg9eRf48rJTB2W)omP#l-BB>hbkU!D<e1MOq+yUbK*{FW%oICD({Bj!Web!xFP1jELSXk4UW+U;;u^SXD)`+*Jv?(+b-9(pShv(4>UJE#Cp<W@DRjzj~<J1XTA6-ZJmXNjs(okUuOI9Mg;kbG{j6F5#Z=cGoA1jf}`2U;XeB8v+;*b!i+SJ#qDxgEC_54@dsk=^B&ZuESBV1$hP6G|uZ+L0AQ+ao&JP`w{SV1njfp-;urekBZmqt^7-~^uO}bH_07YnM_(iUS#nuV*TjPA(_Hm$c&4#-W%FXwR3&mNg~TwmdTo<Qz7kdPKF1eg_?!--bON^!@KTrbR`rKZE+G4NIu!#)eY5%`>+(%E|u}5Ap|&4N%*dasR}|e>49`|h!WrYJS8*cLd~+;WMS232NaPZ$<U1~!ZeZuh;|(-hZh+`I8&g#2C$ycUQyNNUKTi%DUoR;S#Q-xkc}XACyFx4=&<;Eo486IWXzIs8Ai7xc+)uBCZ9skFO;E9Fh>#WM-#kJy9AL@?6!2GxLd$m`c&sezUVP#&bB0h28%@~?_J|%91PB262`=&XkY^J5kVJdc=k4Nr}^9w4In_7a`b(Rj}Mw;DctGQ(??Y?X2)=7y-GT_WpF189)pPz<V{w1vI-+Uh3Da^91q6J)6<b0h9@$-ST2vxR^uVboj4v}oS)0p>B({&$J8VqoE@JG&%*HJB8-peoF22^;Qdu=$aox`A78}FQ5cPf^5pbvd2%LCM$7ZT^88euN26n^AIX#D@o;pqyok?EqO)^3j+TS7;gAZPr!*3!op<Is9V}&d8l6WM=cBXX+45|8JPywXrx)S**?^eI=|yyYA>~*e$B<5=QM^2k&k2@`@o9Vv@c;VyT^lFEvv_oT5iSSOc)W^^k1xi<_+mL6oJ6bf>Ozjsk59uBnm0Zfo-a?&#=}80q5??7h-7;7^a9K=2T825T_GSFg_4`r#wVmB#U-69t^Ei$I$08OLLxC=35M!EgW|!J7liTQ`FPBg#EXb;fSfp}%qTk|xpGCShFEU0=(ZHh0F?mLQ!6=>@0Obc<a3FZOMgj%e;h_<(enJ92Iq-CZVS1>w8MB6T`bQ+RIDUUh@KMTLaI?aI_7-9!|Fy)5s`>80B{MNg0h}CHUK&5oGML)EJYc0V+T&OA<<IvF!u_$g3GoNk|_RB7!Cm}=F=IxFq>55Ju&b|ZQf4R%iVO~i#H?$+1g-WSRuD85jU=0Oc9BV(rP-O!(A)|m@v5M26px)bxU{HUoAZk@_4y>>~Yx<3PV*`%FTv2b_undy8J>N`wY)7%2z~U&O%4R8VjBOm{=o;=rZmP#8Pe&xxxa`kCwIBgn3EWtTMuRg2+QTx)ISP$#do=Vm9zEF3xC``4K-YsKp~~NIYdB1fh`j0T3^=Wb=hMe?QT>hA5IdqR3qi5yHDqC^%yKljw}74BSl<k1zhZBkG<A7M{dJv)f(n-cMZuw$y9&hVr;4?nkPSV6}^RSW=ne>A=%!jlj`t)V53LawRP=++u;VL2uxbmMmT;57n;FWR_z0slzR4xCaknV|i%>xYlafj<k&s<z;5ZYdKs-K1nX-`#mkTJfXE=V#z~-YA|4GBHPMh;hLnCtb_^44{C-?xuW?&TRKWoP$^#2VMtQY9GLtJhTkUHCQ5$kA~Y(At3)zT1_+-SwmA~ES{)9Veg>W<P+GoN7K+ME5gS6^MqKpvj2&=Fq@Q3eHU<teIY${a)M)<9%8=<tVm_Zg)2S_-5yPo4IuoXREH8^`NBUBs9~NtP{5|h-IkjSTY<OWNZF)7pYMTnNsTL*{Glz&cL^#R2obmX#J?EZOxch=wl6$A!!ae0xG_J%2ptiGv=Fgs5H%6ZbLYCU+Lat>>8j@B^W)8I2Xp^N<+*DOwPL7VgZc`vc_Rs$e!$b71_|0g~;#*bcOA;KhIAA$qU~6}bF6|wbT%<^d`=d!Mj(WEAsabw4X^bdk1;-LQjZo<|QiQu3e$(H^{Y|(_X`$uOItw@b@$hWWCjnpeQ%RE+w|z1Y2|?7pc>LpdxLTdd*b+5IBzTT4g`xrmOYi^Tg;_Yz&I0j-scjjV!aSq>rHx116Va;BE{TIMfb7`4&FB?*4&m8KFRQ&B;-9uk(^N@C;<Po1_K%nr2r-M(zt5y>3aLybaZ3GN=<Pibu3#s?Z1@CWsAaeJ>&Vs#bJn@yfD`L3BptPzLi0OSKTngDQf_?tkd##uby_{xS!RlNa6=?r(G;_Wb0Xxidaw(03!9gkMy;}WTAEpui-)=q_mnV~s;nrduG0gHQ~_ENB5CDva33;+Js`+N>+&_Y4{H{L9Drv%@l5Yylfd|vL^8YTZA)D{?dzlucEYJ#5fu?Y48YyXUnK29lW(tWd_)<Ye}A6tXEz$UdiVZ^>-Y5g=B@aTH`o9A>iV4!{TCu6Z6jrmgM?d8MIoe8KsXWcs7a@Q^;+(>kmlakQzgHa6UjAOqMU@x>WKJ1h|jdhvdy<<P12syR$vOBP$7#7VH!dSoBAN0ZuPLK{*`P>5P^oM<`z(uu7A?t*Un7?_)98a9k$1hysricS|zF1sP|k~Yi4X}e|{Of?&0^{JpB`bY#L&KY8bFnXZ$S@MsRPviHY?)14ji5r&BRdVS8uT6b*F|V%e(j`>7b~LPWOf!VA-R@Z2$T$)a$jJr~EK(M6IPOwVR?4o#bKQ_68_Q#c&#jUR=Tn<v)rb=V%ANU44?6{H56;p{-ZZa`rBgMn$#g!T=#5TyWZ*O`%U3)eg6{s{~RMb|%BI$|8@M3ekpx8@CWYt!XV%&78izx-bJD<h&iR2R`q#{-dmlh-dg0{$2+JL$SOk4O-Vh&8S!+`{(VQZ)w^W(&_(#~4L=*26hHylPk>aqPbAAq=~=vu}QT=1m9>^>ZSuc9M^knsUsEI79JVsAQZqC*z<YS%3&|0X7v6x>QBXET^1qqDzSIbU}lhs+od0bz}l`t&Afe!7X!BH>;2Y+q9*3Y|acZOuN>+s!GtHRcT&TB^fn#rFpe*l;3bxr&=trr367}b#g1%Q1zPobUulan}S5;S!3afs4Ht&g%ouv4VJvRKWg{7D>OGlL`#T8rYl%a71L5H>i{j-a$J_Q(<JjcxmrOF|IZUrDhVx|1bzqenXFc139D}HE}yU^oXJa6jhyE`Gd)C7!S={PM+>A}eCnilA;Sj5q7rNX9yR^GW@<>cF@;kx(xop(PJ0xc1)#&8=-`662>*oBudxo@RfP*Jj8p04Fi5TJp38cU4!)Zt9F;z&lSE)WBeldNh4erpy5N)K*H4n)IfUBd;rSg|K|N4{x0ggmC!DkMI>{Mp`|!rGN!YXVPLfT#nZ=4|KRfdvo1XX(5r%!9!OKxR>SpO##~C|}Hpe(RcsN~n;yW>v$6X4XwD^L~7^<?WNtP}qoHA4(EvC-#an&K`oTy^!)H!F4%1GdXJAUC`cqjNT*Q`R0R3_^!KhQkt{S%_p?Yj+0R58z=D#p)x!IZ#ri^#)G0QH>HGi}I}H-)i2Rk+{Rwsp{TI?M0<6q3Z}urWVv!v{WScuZ0PVR2vpE3;+F*)GE-cB`a`x&<*r&gNI^`S6oHK3eD#@((1(X5)poU>$^71wk5aNkQYQ6hW{JlQak%O(p6o!ad&?$X%wiMt`0hFGzC2BJSRO>16S8BdxOe+uJxHH3UXPklm`ttQKtGb0p_l*}<(eUk;;UER_}f1*^oXF-g?9M;4oBh%C-HW2nizS0V=sV`kgjgrV6rr$de$tFrFB2emP$$6hn96FOYH>SV?;`q9=;2wTC9s5X{J&E~vIfYLqWD{Xt&h7B%`IP*CviI-Hr+U&|3vl?aBFlMm~M<p=xz?Si1`~-~!Q>#c5?FM#X<O2_}2nm3jt*nyBBa0@)C6zK0L_uT^z?Lw6T(VcHlAW)l^)JI^!eJ-~4wR!6(J@S-C@hME2G;&k+y%Mw>X0u|D+CV55v$dVx5nS50;j(C7v;M#F6}xEp%lukG*6e0LRkgu2F})N{)@o5c2ipU+Prk?yG$yc4}V>D6iB}Cqb)3$UsoL8o*W87i+;bc-M$vxcK5>4-LNyw)V988DY=a3aKSBjUxc}_!gM_&K-(KFv*k)v1SRXernetUxO7}SEOY~fH|3`HGcD*qH34+S`Lmi~8yC42sYHl?Qzuzs9-mVd-5_LXJFF+?I+HRH5RPc%=hn!j+^qb7hQ-)CEM{lBo;mPcSd=a=4q(X#8R0D_iRxl#J0m493jj+;OSp<@C}wH%RD~uQyFkY#)ENp`>`>#GQ;)J}6zd>J(xeIk7vO?N*f;8SbRfult=lYNq+(fMxWr|wam?6y96~F{CBg^#4<5W8t`DK0SgL3Mown(400q4q>4y)~K`Z&510%G{&{Q#N1*r<FYDi$@0~nhqF7rzp*^SoCGS9RkJh)$js)z7QIU*=54M1HSdvGE8YYotSBzFG*^^br+_9x;gRwsiYW8*+g<+b!`FbUD%wluA$3%C03ojM6T5hQTzStDx!UUaveO*@S_b<%JunrPL8`J1@Ek;IY;W?&rRVW|2l?8F)%i|ta$gBJHP=l2P+$+HbvO}qeTJoKu!<UUNRGWbAckbAn2XEFPN8)O4rp>zcl_@;za1}Yi)&E%L`PyAij*xzb;pKGc7f0Q=g8u0U$?gvtRYtd;}`KDe>JeMkxSMPL5dL-#I=*(d~8cIQ_1iaPmsBXvpMyd5uj_+B7(=|jCrk|^$)9e8ixUSS8LFR8YU3z<JGnGy{`-HP$m5^rN>BbgFXDgAZXO*-&>Aieg^XzXww9p6G^W7){KOrTdA4{A=4h^#uM7WRBI%;8Ge${n3Vj7D`yAdJMNSv~X_Lfy4VIZ*}AQ?@{KHu<OkqEp|@w&vEC<6E>KPpE@A6Pw5WDJG92eUw1O~PcDXw@8Aq{a%=0~}9FuKQ2BHPtwO`(^L{UAHBBEuWP9m*Q7ElwbEiDvo(ABYiI0-02W^t67`H*L=29SpMNH30p#^VNjYo=@#K-&8L2aNvVDv5k`IL7F1n>Q41L&qC!O=8X19c6=AwpAW`H+r7^B6$UMmFYcv^5|10HAgj`pY<6rY1mwI?w$81DCBG$ZyV(st9sv>(;T{@r{S|um#5~7Whf#EHn;ZR>HCg9Y9tcBy-1z$h)z-t>(0IXphhXx>M)B;`HoAFO*G<G60P87e*i5*)&`xmq_zq7UzTHI(b@ah4Toi*TGTT6){yw%Rm_Jwj(VV9IaY3}6EM!`N+{ub?_*Y^bAK{2FmC;!*J6m#8lVcp@8O6PB`f({157IjJXAd~iar98vemQR=}sRZF}3|sJsK~LvYn;o;<Ww_H3XiQG|`hSP~&XiM!jbIix7`k?VC=`xt{PBoDFOS$6?SeLj1$I4_wJySiLA<*Ef(!G!DyL@qR0N9>JrOulfnll*z^`E#1m!6BptwQ9moZSY&?mF6I-}(QEX@PMsMUSZizaHAfdCK|#2QR=&oY^*2~bp*pn%MEnCkm(36qPcVt-mp+L2>hTDH;@wXpoEWE6Mhfw<np<Q9^I>epz1*tHWE56Z48*-9nB{!dD--V^b+-z>^>_0w^?ugkjZR;kq$tR&#n$BzpR-|&czi^$>-=gp1WvJ>OEVs4@tQ^SpFxVqsP%}#nkvzMh6cd857&RhC<mb*I0WP5$}g-1T)!~)SxhuW`v<IP(r5licCf;;94IwJ1MC&Ks2`MrkM7{fNd;E!=wy-f5)m)4Bes(@3ntm=>sf4M}zELz?yT&z+%`Un4Xg;WUqb;KVySpDr&dgij;ZmWaweCBAhvBYRt&#XVhJs6sQTSBl~iOBkJzxmgR_(Ke(mf(T~I2@_7OA4~w#3o4;nZIHwt9v8^Hw2G+`-a?m{92<Gh*nD>Sh8Eg#Rv}bofC^H30sxc9vt5K4j=QLaH2LG@H6K~PXv$eVC{T6>K^xYpJ0rgJnI&J)Y&-=YrHn;RkD?iGeOcjaXh3)9B`poj1n0QYy{*cyFytMaN~t>Oq|ozJab3p!w%`KhtnfW;?Vb;eyF}B^;GK63-LXPh8O#MmN+B(u=Wz)_=W-8uZ@?i6T3_Um1{VeTLDAyoi=9WL+-{!vZ^>PL#7RBQhDL8aZAP>_}FA0`k(T$ffu-H7QzN!i_~g0UX`_I-ju;<-d56X$;LEsG&QPFZjf(;J&*)cP~a!T&k7t<4sq_PA*-81c6*k?{o#M_+AZCSU^iT+Wxi{FVbpR~;`2PKB{<w_d8yt)ehW2h9KvjQAq`&5Bof{OTC-Q<xYkZ-y{Icp9WC9|0Sl-LO&uY`)G=;Qxf+;#l`W{7&?Dhw+8Dgk+UT9)C&1NMGs5b3kI<X%>tnp@Q}((77gK{2-sGvqgYy6Ah4YqA^-4KDJXM=j(=LAziDe<d62Xa4e><yj8b2tXF_|9b#Mw?`j>*(-5uAPX8%?_M4s~*LlGd(}eSQ7N<O~xF(~8_*-o3y6x7V-!^8VFJarNfye?hi@+msFaLBe^BgD~0X7aTL(e^WHCa@a4$0oF86-xSF@Nl7~VjhEQI#=$Vp;iV4u^4mt8{I(*}+}z5swzAyJZ6w#e_+(!pQg@ngt;*zG+w5KimhNwG6G?}D62<Oul|_W4eReJeEFT-6Djeu4Amo~_Xz~33{!|$sGERk;1j*JU!>dGixPx%z;#LQX2n*)wK)IJCFX9XVqK`<GFQJ4=>zW2hCyw9<7oR*TJK66Hi%QQP{FX@3w_J!M0snWp;Mm&>HC@hAz7oKz!W~C-zmd$CtKWLZ@m)pj!-CL|EQ_#sRKdD=`XB7fJ9tf^SEJVUt@G`QH4zfuwcxBay;1i!Qe8c0-|`^Zf!<ZBs~4mhs5=|PZ9q8uNd~nT9&R8q`1S7-zR-~Icw76D0`h9CdkP(Wz{-EzH&c4z^$-8~XSgZezxf%U6C8H<p?e|bf7I)4`OoA1$dGm4g~x-Fy|TURc{^1oKJ25%co8R{4LYeHhh6sE;oH3Ss%g0ATQ^|Wo_qov&M89q0pM_xMY-PLr@}>>9{bdDnnt~uW(UIn>Pn75QYUQ1nVtCj;t=`@fr0!vZ*kcPQh{Hox)w$`vw#;f>83ETYO;qj4RIH9rCwJTZ+jsLmk0j7S#WnG)!;?bU~)h^XpoB5mmM@UuHiirY3C8vsdJnZ*U(X&#X9<GW)^oHboN7ql8Dj`RQw$q%e@FqB{ab(eG975Bumc_6Xz-QtUR^`W$hOE4R3h|4$Px>K}T@F7J5vDtlHuL^bS^p2O^jXY$cuR%zsNZt1G#kOQ^(`58GJHrKiT+I`G-t=8D1VH!oiu$S<F#sL&Tr#4oVTf8nczmvz?pOnDHYD=nIgb97UwKbI=0M_*08RafJzEc7SiHUIw-(CtVL-@h>+vv~eWa>wru@m#8nomwed3sPKuey6`Y1#iT}h&6m8XJR?9^miW_xGysdq`y$d7sGWLwxM552DYo5rhzZ3cKZBnxn`ru$yajCTwbU3Ipn`7;cW1KI@gT1M`fFl9d@%zZ`mF^{=1}}0nhL!?dP6-`ER%PnN|"},"regeneration":["git clone https://github.com/gtmcknight/amm-sim src && git -C src checkout 4812a6382311624e1506aca131ca35c378f930d9; (cd src/research/solved && python3 -I gates.py)","in w/: combo.py, twocomp.py from the COMBO/TC artifacts' Code blocks; blob.bin, qmref.py from identity-md-launches/goal-3170-beat-network-s-amm-entry@31ff8ed1 reference/qt1/; decode the bundle","python3 -I mkri.py ../src/research/solved/replica2.py replica2i.py; python3 -I instcheck.py; python3 -I gq.py","for m in combo pp qt1: python3 -I screen.py $m 64000 65024 6; python3 -I diag.py 64000 64256 6","for m in combo pp qt1: python3 -I screen.py $m 10000 12048 6   (once, after freezing sha256s)","python3 -I agg.py result.json"]}