{ "blocked": null, "sources": [ { "url": "https://www.linkedin.com/pulse/amm-fee-challenge-notes-from-our-run-optimization-arena-qe3ff", "author": "Algebra Finance / Algebra DEX engine", "date": "2026-03-27", "holds": "First-hand entrant article: peak ninth among 654; twelfth when written. Describes environment estimation, short-term response, directional fee protection and probabilistic block ordering. No numerical formula, constants, ablation scores or source code published in this article." }, { "url": "https://www.benedict.dev/optimization-arena-learnings", "author": "Benedict Brady (organizer; not an entrant disclosure)", "date": "2026-04-09", "holds": "Talk-derived organizer account of the top eight: two Solidity neural networks; compact 200–300-line mathematical models; approaches with over 100 constants; one competitor-AMM model. Reports near-identical outcomes across 20 simulations and synchronized protective-side flips. Does not identify which entrant used which architecture or disclose weights/formulas." }, { "url": "https://raw.githubusercontent.com/jiayaoqijia/amm-challenge-yq/3e9dc7d5a6192c88b562a0f8d9c172ca566ec94d/README.md", "author": "jiayaoqijia", "date": null, "holds": "Entrant repository write-up. Claims official leaderboard score 523.21; local benchmark 495 simulations over five seed offsets. Describes adaptive shock gate, cubic toxicity, trade-aligned boost, asymmetric stale discount and tail compression. No claimed board rank. Snapshot commit 3e9dc7d5a6192c88b562a0f8d9c172ca566ec94d." }, { "url": "https://raw.githubusercontent.com/dipankar/amm-challenge/c588ae6c47fc9af2063c9bb1f2efbeb0dd276dab/ANALYSIS.md", "author": "dipankar", "date": null, "holds": "First-hand research post-mortem. Claims SigmaAdaptive 526.54 versus WinningV5 295.57; attributes the change to state estimation instead of arb/retail classification. Gives an adaptive price gate, cubic toxicity and aligned boost; cites jiayaoqijia as reference. These are author claims, not fresh board measurements." }, { "url": "https://raw.githubusercontent.com/dipankar/amm-challenge/c588ae6c47fc9af2063c9bb1f2efbeb0dd276dab/README.md", "author": "dipankar", "date": null, "holds": "Labels 526.54 as a 500-simulation local edge, not an official board score. Snapshot commit c588ae6c47fc9af2063c9bb1f2efbeb0dd276dab." }, { "url": "https://raw.githubusercontent.com/ChengHua926/amm-challenge/d34a5199c59a4c44d5336a25f1436b4ae1f21104/contracts/src/Strategy.sol", "author": "ChengHua926", "date": null, "holds": "Pinned public Solidity reference supplied by the task, getName no_cubic. Code evidence only; no accompanying first-hand narrative or current board claim established here." }, { "url": "https://raw.githubusercontent.com/jiayaoqijia/amm-challenge-yq/3e9dc7d5a6192c88b562a0f8d9c172ca566ec94d/contracts/src/Strategy.sol", "author": "jiayaoqijia", "date": null, "holds": "Exact Solidity accompanying the YQ write-up; validated and measured without modification." }, { "url": "https://raw.githubusercontent.com/dipankar/amm-challenge/c588ae6c47fc9af2063c9bb1f2efbeb0dd276dab/contracts/src/SigmaAdaptiveStrategy.sol", "author": "dipankar", "date": null, "holds": "Exact Solidity accompanying the post-mortem; validated and measured without modification. Its attract ratio is 1.12; the analysis table describes 1.124 for the cited reference, so they are not identical policies." }, { "url": "https://api.imd.fun/artifacts/19f2ec857df536e8ab62ab1da89ec6623332521ce46849f15bccff402c63e1e5", "author": "IdentityMD (task-specified comparator)", "date": null, "holds": "RS2 source, SHA-256 f24b6ca198b744c380842c3a17964a6dc21f311bb04984b26733afa3fee8eaf2, verified from decoded code bytes." }, { "url": "https://api.imd.fun/artifacts/33f54e81f3372088e9eced52c0174f310ac348af98ef5727b73785c088679194", "author": "IdentityMD (task-specified measurement reference)", "date": null, "holds": "Supplied harness_source describes independent candidate-versus-Vanilla arms, pairing and retrospective arb/retail decomposition. Inspected as data; a local measurement program was written for this item." }, { "url": "https://raw.githubusercontent.com/benedictbrady/amm-challenge/6b6921f8d463314aa7c96234e1be25a25245f704/amm_competition/competition/match.py", "author": "benedictbrady/amm-challenge", "date": null, "holds": "MatchRunner._build_configs draws retail size, rate and volatility in that order using default_rng(seed)." }, { "url": "https://raw.githubusercontent.com/benedictbrady/amm-challenge/6b6921f8d463314aa7c96234e1be25a25245f704/amm_sim_rs/src/simulation/engine.rs", "author": "benedictbrady/amm-challenge", "date": null, "holds": "Pinned engine seeds price with seed and retail with seed+1; arbitrage precedes retail." }, { "url": "https://raw.githubusercontent.com/benedictbrady/amm-challenge/6b6921f8d463314aa7c96234e1be25a25245f704/amm_sim_rs/src/market/arbitrageur.rs", "author": "benedictbrady/amm-challenge", "date": null, "holds": "Closed-form arbitrage size/profit formulas used for retrospective edge decomposition, matching the supplied harness." } ], "mechanisms": [ { "name": "Algebra: two adaptation layers", "rule": "Maintain estimates of volatility, retail flow, arbitrage probability and normalizer behavior from trade/pool history; also react to immediate pool price movements. Formula, state layout and constants not disclosed.", "quote": "two layers of adaptation", "url": "https://www.linkedin.com/pulse/amm-fee-challenge-notes-from-our-run-optimization-arena-qe3ff", "board": "Author claims peak #9/654; #12 at publication" }, { "name": "Algebra: asymmetric protection", "rule": "Increase the fee toward the external price when retail displaces pool price, reducing arbitrage extraction while retaining retail. Numerical mapping and constants not disclosed.", "quote": "importance of asymmetric fees", "url": "https://www.linkedin.com/pulse/amm-fee-challenge-notes-from-our-run-optimization-arena-qe3ff", "board": "Same submission; peak #9/654 claimed" }, { "name": "Algebra: within-block probability", "rule": "Use probabilistic block evolution because arbitrage can only be first and later retail count is unknown. No probability model or constants disclosed.", "quote": "probabilistic estimates", "url": "https://www.linkedin.com/pulse/amm-fee-challenge-notes-from-our-run-optimization-arena-qe3ff", "board": "Same submission; peak #9/654 claimed" }, { "name": "Top-eight architecture diversity (organizer evidence)", "rule": "Brady reports neural-network, compact mathematical, many-constant and competitor-model implementations. No identities, exact policies, network architectures or weights disclosed; not executable rules.", "quote": "two black box neural networks", "url": "https://www.benedict.dev/optimization-arena-learnings", "board": "Top eight at time of April 9 article; no individual ranks or scores" }, { "name": "Top-eight shared signals (organizer evidence)", "rule": "Estimate arb probability, skew using arb information, adjust widths to estimated simulation volatility and suppress outlier-induced skew. No formulas or constants disclosed.", "quote": "estimating probabilities of arb trades", "url": "https://www.benedict.dev/optimization-arena-learnings", "board": "Top eight collectively; no entrant mapping" }, { "name": "YQ: adaptive shock gate and asymmetric toxicity", "rule": "Write-up: 11 state slots (fees, timestamp, direction, activity, pHat, sigmaHat, lambdaHat, sizeHat, toxEma, count); alpha_first=0.26; toxicity includes a cubic term; aligned trades add protection; attract stale discount exceeds protect shift; tail slopes protect=0.93, attract=0.955. Numerical gate, cubic coefficient and full arithmetic are in its cited Solidity, summarized separately below.", "quote": "Adaptive Shock Gate", "url": "https://raw.githubusercontent.com/jiayaoqijia/amm-challenge-yq/3e9dc7d5a6192c88b562a0f8d9c172ca566ec94d/README.md", "board": "Official score 523.21 claimed; no rank" }, { "name": "YQ: implementation constants", "rule": "WAD arithmetic: pImplied=spot*(1-feeUsed) for isBuy, else spot/(1-feeUsed); update pHat with alpha=0.26 first-in-step, 0.05 later if relative return<=max(10*sigmaHat,0.03). Cubic toxicity coefficient 15000 BPS; aligned boost 2500 BPS*tradeRatio; stale shift 6850 BPS*toxSignal, attract multiplier 1.124; base 3 BPS; tail knee 500 BPS, slopes 0.93/0.955. These are selected source components, not a replacement for the full contract.", "quote": "uint256 constant PHAT_ALPHA = 260000000000000000;", "url": "https://raw.githubusercontent.com/jiayaoqijia/amm-challenge-yq/3e9dc7d5a6192c88b562a0f8d9c172ca566ec94d/contracts/src/Strategy.sol", "board": null }, { "name": "SigmaAdaptive post-mortem: estimate state rather than classify trades", "rule": "Author gives toxicity=abs(spot-pHat)/pHat; gate=max(10*sigmaHat,3%); pHat alpha=0.26 on first-in-step and 0.05 subsequently; cubic term=15000 bps*tox^3; aligned boost=2500 bps*tradeRatio. State: pHat, sigmaHat, tox, dir, act. Reports pHat-removal ablation -0.42 and other features individually within noise; not independently reproduced as ablations here.", "quote": "The Winning Paradigm: State Estimation", "url": "https://raw.githubusercontent.com/dipankar/amm-challenge/c588ae6c47fc9af2063c9bb1f2efbeb0dd276dab/ANALYSIS.md", "board": "526.54 local claim, not verified leaderboard rank/score" }, { "name": "CH reference: gated implied-price tracking", "rule": "State slots 0..10: bid,ask,lastTs,dirState,actEma,pHat,sigmaHat,lambdaHat,sizeHat,toxEma,stepTradeCount. pImplied=spot*(1-feeUsed) for isBuy, else spot/(1-feeUsed); pHat blend alpha=0.27 first-in-step or 0.06 later, gated by max(7.2*sigmaHat,0.03). Initial fee=4 BPS; sigmaHat=0.00095; lambdaHat=0.8; sizeHat=0.002. Source has quadratic toxicity, no cubic term. Full source is included if this is the best measured file.", "quote": "uint256 constant PHAT_ALPHA_RETAIL = 60000000000000000;", "url": "https://raw.githubusercontent.com/ChengHua926/amm-challenge/d34a5199c59a4c44d5336a25f1436b4ae1f21104/contracts/src/Strategy.sol", "board": null } ], "files": [ { "url": "https://raw.githubusercontent.com/dipankar/amm-challenge/c588ae6c47fc9af2063c9bb1f2efbeb0dd276dab/contracts/src/SigmaAdaptiveStrategy.sol", "sha256": "84da373d4ecc47b5ecaa96adfcc432ad62825a0c608e7529404fa8a0184cdf90", "valid": true, "d32": -8.01644894093358, "oog": 0, "screen": { "n": 32, "mean_diff": -8.01644894093358, "se": 0.515870965236206, "worst": [ { "seed": 900025, "diff": -17.895737224713457 }, { "seed": 900012, "diff": -13.059750628236316 }, { "seed": 900010, "diff": -12.752061965009034 }, { "seed": 900005, "diff": -11.942326408590702 }, { "seed": 900021, "diff": -11.046649032753635 } ] }, "gas32": { "max_ok": 148213, "oog": 0, "init_max": 176674, "reverts": 0, "other_halts": 0, "tx_errors": 0, "bad_returns": 0 } }, { "url": "https://raw.githubusercontent.com/ChengHua926/amm-challenge/d34a5199c59a4c44d5336a25f1436b4ae1f21104/contracts/src/Strategy.sol", "sha256": "6dd68f77a70900319460215374143dbd5d924198450a46d9f5f91f6b1e3d35c3", "valid": true, "d32": -1.9468584042492232, "oog": 0, "screen": { "n": 32, "mean_diff": -1.9468584042492232, "se": 0.20503310917898884, "worst": [ { "seed": 900014, "diff": -4.306222491488484 }, { "seed": 900012, "diff": -3.7863019326727 }, { "seed": 900017, "diff": -3.763822195092075 }, { "seed": 900023, "diff": -3.572897158206729 }, { "seed": 900010, "diff": -3.3486129067644015 } ] }, "gas32": { "max_ok": 128929, "oog": 0, "init_max": 185532, "reverts": 0, "other_halts": 0, "tx_errors": 0, "bad_returns": 0 } }, { "url": "https://raw.githubusercontent.com/jiayaoqijia/amm-challenge-yq/3e9dc7d5a6192c88b562a0f8d9c172ca566ec94d/contracts/src/Strategy.sol", "sha256": "00fae418c49cd681ad0879c6f96ebed0163fab5a660b38030f11e6947a39afd4", "valid": true, "d32": -2.8743337699932106, "oog": 0, "screen": { "n": 32, "mean_diff": -2.8743337699932106, "se": 0.24043747036278731, "worst": [ { "seed": 900005, "diff": -5.428194593183548 }, { "seed": 900022, "diff": -5.225871391006194 }, { "seed": 900017, "diff": -4.878984596833789 }, { "seed": 900010, "diff": -4.691139704285888 }, { "seed": 900025, "diff": -4.598732433260352 } ] }, "gas32": { "max_ok": 127893, "oog": 0, "init_max": 185532, "reverts": 0, "other_halts": 0, "tx_errors": 0, "bad_returns": 0 } } ], "top3": [ { "sha256": "6dd68f77a70900319460215374143dbd5d924198450a46d9f5f91f6b1e3d35c3", "url": "https://raw.githubusercontent.com/ChengHua926/amm-challenge/d34a5199c59a4c44d5336a25f1436b4ae1f21104/contracts/src/Strategy.sol", "paired": { "n": 256, "mean_diff": -1.8404066162914532, "se": 0.06979100904354843, "worst": [ { "seed": 901156, "diff": -8.517896528998108 }, { "seed": 901180, "diff": -5.985494457703794 }, { "seed": 901048, "diff": -4.914895708905533 }, { "seed": 901142, "diff": -4.7297253625318945 }, { "seed": 901033, "diff": -4.402695654344257 } ] }, "retail_edge_delta": -0.008515070775243494, "arb_edge_delta": -1.831891545516214, "gas": { "max_ok": 151576, "oog": 0, "init_max": 185532, "reverts": 0, "other_halts": 0, "tx_errors": 0, "bad_returns": 0 }, "mechanism": "CH reference: gated implied-price tracking", "valid": true, "mean_edge": 531.8947526230002 }, { "sha256": "00fae418c49cd681ad0879c6f96ebed0163fab5a660b38030f11e6947a39afd4", "url": "https://raw.githubusercontent.com/jiayaoqijia/amm-challenge-yq/3e9dc7d5a6192c88b562a0f8d9c172ca566ec94d/contracts/src/Strategy.sol", "paired": { "n": 256, "mean_diff": -2.730905464885787, "se": 0.07757919730518127, "worst": [ { "seed": 901156, "diff": -9.29795016706089 }, { "seed": 901180, "diff": -8.009805892983081 }, { "seed": 901133, "diff": -5.570670043617838 }, { "seed": 901204, "diff": -5.373609550495189 }, { "seed": 901048, "diff": -5.3075022238934935 } ] }, "retail_edge_delta": -1.8372431375745026, "arb_edge_delta": -0.8936623273112868, "gas": { "max_ok": 147039, "oog": 0, "init_max": 185532, "reverts": 0, "other_halts": 0, "tx_errors": 0, "bad_returns": 0 }, "mechanism": "YQ: adaptive shock gate and asymmetric toxicity", "valid": true, "mean_edge": 531.0042537744058 }, { "sha256": "84da373d4ecc47b5ecaa96adfcc432ad62825a0c608e7529404fa8a0184cdf90", "url": "https://raw.githubusercontent.com/dipankar/amm-challenge/c588ae6c47fc9af2063c9bb1f2efbeb0dd276dab/contracts/src/SigmaAdaptiveStrategy.sol", "paired": { "n": 256, "mean_diff": -7.964970098691098, "se": 0.2045763429268462, "worst": [ { "seed": 901180, "diff": -43.041805616406464 }, { "seed": 901063, "diff": -31.917152397542736 }, { "seed": 901033, "diff": -14.532722939904147 }, { "seed": 901156, "diff": -13.981884028697664 }, { "seed": 901048, "diff": -13.009996739116445 } ] }, "retail_edge_delta": -0.7191316719923646, "arb_edge_delta": -7.245838426698733, "gas": { "max_ok": 172035, "oog": 0, "init_max": 176674, "reverts": 0, "other_halts": 0, "tx_errors": 0, "bad_returns": 0 }, "mechanism": "SigmaAdaptive post-mortem: estimate state rather than classify trades", "valid": true, "mean_edge": 525.7701891406006 } ], "code": "// SPDX-License-Identifier: MIT\npragma solidity ^0.8.24;\n\nimport {AMMStrategyBase} from \"./AMMStrategyBase.sol\";\nimport {TradeInfo} from \"./IAMMStrategy.sol\";\n\ncontract Strategy is AMMStrategyBase {\n uint256 constant ELAPSED_CAP = 8;\n uint256 constant SIGNAL_THRESHOLD = WAD / 650;\n uint256 constant DIR_DECAY = 750000000000000000;\n uint256 constant ACT_DECAY = 530000000000000000;\n uint256 constant SIZE_DECAY = 530000000000000000;\n uint256 constant TOX_DECAY = 910000000000000000;\n uint256 constant SIGMA_DECAY = 690000000000000000;\n uint256 constant LAMBDA_DECAY = 990000000000000000;\n uint256 constant SIZE_BLEND_DECAY = 828000000000000000;\n uint256 constant TOX_BLEND_DECAY = 51000000000000000;\n uint256 constant ACT_BLEND_DECAY = 985000000000000000;\n uint256 constant PHAT_ALPHA = 270000000000000000;\n uint256 constant PHAT_ALPHA_RETAIL = 60000000000000000;\n uint256 constant DIR_IMPACT_MULT = 4;\n uint256 constant GATE_SIGMA_MULT = 7200000000000000000;\n uint256 constant MIN_GATE = 30000000000000000;\n uint256 constant TRADE_TOX_BOOST = 3500 * BPS;\n uint256 constant STALE_ATTRACT_FRAC = 1050000000000000000;\n uint256 constant RET_CAP = WAD / 35;\n uint256 constant TOX_CAP = WAD / 5;\n uint256 constant TRADE_RATIO_CAP = WAD / 5;\n uint256 constant LAMBDA_CAP = 1820000000000000000;\n uint256 constant STEP_COUNT_CAP = 64;\n uint256 constant BASE_FEE = 4 * BPS;\n uint256 constant SIGMA_COEF = 200000000000000000;\n uint256 constant LAMBDA_COEF = 12 * BPS;\n uint256 constant FLOW_SIZE_COEF = 4842 * BPS;\n uint256 constant TOX_COEF = 300 * BPS;\n uint256 constant TOX_QUAD_COEF = 15000 * BPS;\n uint256 constant ACT_COEF = 91843 * BPS;\n uint256 constant DIR_COEF = 20 * BPS;\n uint256 constant DIR_TOX_COEF = 100 * BPS;\n uint256 constant STALE_DIR_COEF = 6850 * BPS;\n uint256 constant SIGMA_TOX_COEF = 500 * BPS;\n uint256 constant TAIL_KNEE = 600 * BPS;\n uint256 constant TAIL_SLOPE_PROTECT = 930000000000000000;\n uint256 constant TAIL_SLOPE_ATTRACT = 850000000000000000;\n\n function afterInitialize(uint256 initialX, uint256 initialY) external override returns (uint256, uint256) {\n slots[0] = BASE_FEE; slots[1] = BASE_FEE; slots[2] = 0; slots[3] = WAD; slots[4] = 0;\n slots[5] = initialX > 0 ? wdiv(initialY, initialX) : 100 * WAD;\n slots[6] = 950000000000000; slots[7] = 800000000000000000; slots[8] = 2000000000000000;\n slots[9] = 0; slots[10] = 0;\n return (BASE_FEE, BASE_FEE);\n }\n\n function afterSwap(TradeInfo calldata trade) external override returns (uint256, uint256) {\n uint256 prevBidFee = slots[0]; uint256 prevAskFee = slots[1]; uint256 lastTs = slots[2];\n uint256 dirState = slots[3]; uint256 actEma = slots[4]; uint256 pHat = slots[5];\n uint256 sigmaHat = slots[6]; uint256 lambdaHat = slots[7]; uint256 sizeHat = slots[8];\n uint256 toxEma = slots[9]; uint256 stepTradeCount = slots[10];\n\n bool isNewStep = trade.timestamp > lastTs;\n if (isNewStep) {\n uint256 elapsedRaw = trade.timestamp - lastTs;\n uint256 elapsed = elapsedRaw > ELAPSED_CAP ? ELAPSED_CAP : elapsedRaw;\n dirState = _decayCentered(dirState, DIR_DECAY, elapsed);\n actEma = wmul(actEma, _powWad(ACT_DECAY, elapsed));\n sizeHat = wmul(sizeHat, _powWad(SIZE_DECAY, elapsed));\n toxEma = wmul(toxEma, _powWad(TOX_DECAY, elapsed));\n if (stepTradeCount > 0 && elapsedRaw > 0) {\n uint256 lambdaInst = (stepTradeCount * WAD) / elapsedRaw;\n if (lambdaInst > LAMBDA_CAP) lambdaInst = LAMBDA_CAP;\n lambdaHat = wmul(lambdaHat, LAMBDA_DECAY) + wmul(lambdaInst, WAD - LAMBDA_DECAY);\n }\n stepTradeCount = 0;\n }\n\n bool firstInStep = stepTradeCount == 0;\n uint256 spot = trade.reserveX > 0 ? wdiv(trade.reserveY, trade.reserveX) : pHat;\n if (pHat == 0) pHat = spot;\n\n uint256 feeUsed = trade.isBuy ? prevBidFee : prevAskFee;\n uint256 gamma = feeUsed < WAD ? WAD - feeUsed : 0;\n uint256 pImplied;\n if (gamma == 0) { pImplied = spot; }\n else { pImplied = trade.isBuy ? wmul(spot, gamma) : wdiv(spot, gamma); }\n\n {\n uint256 ret = pHat > 0 ? wdiv(absDiff(pImplied, pHat), pHat) : 0;\n uint256 alpha = firstInStep ? PHAT_ALPHA : PHAT_ALPHA_RETAIL;\n uint256 adaptiveGate = wmul(sigmaHat, GATE_SIGMA_MULT);\n if (adaptiveGate < MIN_GATE) adaptiveGate = MIN_GATE;\n if (ret <= adaptiveGate) { pHat = wmul(pHat, WAD - alpha) + wmul(pImplied, alpha); }\n if (firstInStep) {\n if (ret > RET_CAP) ret = RET_CAP;\n sigmaHat = wmul(sigmaHat, SIGMA_DECAY) + wmul(ret, WAD - SIGMA_DECAY);\n }\n }\n\n uint256 tradeRatio = trade.reserveY > 0 ? wdiv(trade.amountY, trade.reserveY) : 0;\n if (tradeRatio > TRADE_RATIO_CAP) tradeRatio = TRADE_RATIO_CAP;\n\n if (tradeRatio > SIGNAL_THRESHOLD) {\n uint256 push = tradeRatio * DIR_IMPACT_MULT;\n if (push > WAD / 4) push = WAD / 4;\n if (trade.isBuy) {\n dirState = dirState + push;\n if (dirState > 2 * WAD) dirState = 2 * WAD;\n } else {\n dirState = dirState > push ? dirState - push : 0;\n }\n actEma = wmul(actEma, ACT_BLEND_DECAY) + wmul(tradeRatio, WAD - ACT_BLEND_DECAY);\n sizeHat = wmul(sizeHat, SIZE_BLEND_DECAY) + wmul(tradeRatio, WAD - SIZE_BLEND_DECAY);\n if (sizeHat > WAD) sizeHat = WAD;\n }\n\n uint256 tox = pHat > 0 ? wdiv(absDiff(spot, pHat), pHat) : 0;\n if (tox > TOX_CAP) tox = TOX_CAP;\n toxEma = wmul(toxEma, TOX_BLEND_DECAY) + wmul(tox, WAD - TOX_BLEND_DECAY);\n uint256 toxSignal = toxEma;\n\n stepTradeCount = stepTradeCount + 1;\n if (stepTradeCount > STEP_COUNT_CAP) stepTradeCount = STEP_COUNT_CAP;\n\n uint256 flowSize = wmul(lambdaHat, sizeHat);\n uint256 fBase = BASE_FEE + wmul(SIGMA_COEF, sigmaHat) + wmul(LAMBDA_COEF, lambdaHat) + wmul(FLOW_SIZE_COEF, flowSize);\n uint256 fMid = fBase + wmul(TOX_COEF, toxSignal) + wmul(TOX_QUAD_COEF, wmul(toxSignal, toxSignal)) + wmul(ACT_COEF, actEma);\n fMid = fMid + wmul(SIGMA_TOX_COEF, wmul(sigmaHat, toxSignal));\n\n uint256 dirDev; bool sellPressure;\n if (dirState >= WAD) { dirDev = dirState - WAD; sellPressure = true; }\n else { dirDev = WAD - dirState; sellPressure = false; }\n\n uint256 skew = wmul(DIR_COEF, dirDev) + wmul(DIR_TOX_COEF, wmul(dirDev, toxSignal));\n uint256 bidFee; uint256 askFee;\n if (sellPressure) {\n bidFee = fMid + skew;\n askFee = fMid > skew ? fMid - skew : 0;\n } else {\n askFee = fMid + skew;\n bidFee = fMid > skew ? fMid - skew : 0;\n }\n\n {\n uint256 staleShift = wmul(STALE_DIR_COEF, toxSignal);\n uint256 attractShift = wmul(staleShift, STALE_ATTRACT_FRAC);\n if (spot >= pHat) {\n bidFee = bidFee + staleShift;\n askFee = askFee > attractShift ? askFee - attractShift : 0;\n } else {\n askFee = askFee + staleShift;\n bidFee = bidFee > attractShift ? bidFee - attractShift : 0;\n }\n }\n\n {\n bool tradeAligned = (trade.isBuy && spot >= pHat) || (!trade.isBuy && spot < pHat);\n if (tradeAligned) {\n uint256 tradeBoost = wmul(TRADE_TOX_BOOST, tradeRatio);\n if (trade.isBuy) { bidFee = bidFee + tradeBoost; }\n else { askFee = askFee + tradeBoost; }\n }\n }\n\n if (sellPressure) {\n bidFee = clampFee(_compressTailWithSlope(bidFee, TAIL_SLOPE_PROTECT));\n askFee = clampFee(_compressTailWithSlope(askFee, TAIL_SLOPE_ATTRACT));\n } else {\n askFee = clampFee(_compressTailWithSlope(askFee, TAIL_SLOPE_PROTECT));\n bidFee = clampFee(_compressTailWithSlope(bidFee, TAIL_SLOPE_ATTRACT));\n }\n\n slots[0] = bidFee; slots[1] = askFee; slots[2] = trade.timestamp;\n slots[3] = dirState; slots[4] = actEma; slots[5] = pHat;\n slots[6] = sigmaHat; slots[7] = lambdaHat; slots[8] = sizeHat;\n slots[9] = toxEma; slots[10] = stepTradeCount;\n\n return (bidFee, askFee);\n }\n\n function _compressTailWithSlope(uint256 fee, uint256 slope) internal pure returns (uint256) {\n if (fee <= TAIL_KNEE) return fee;\n return TAIL_KNEE + wmul(fee - TAIL_KNEE, slope);\n }\n\n function _powWad(uint256 factor, uint256 exp) internal pure returns (uint256 result) {\n result = WAD;\n while (exp > 0) {\n if (exp & 1 == 1) result = wmul(result, factor);\n factor = wmul(factor, factor);\n exp >>= 1;\n }\n }\n\n function _decayCentered(uint256 centered, uint256 decayFactor, uint256 elapsed) internal pure returns (uint256) {\n uint256 mul = _powWad(decayFactor, elapsed);\n if (centered >= WAD) { return WAD + wmul(centered - WAD, mul); }\n uint256 below = wmul(WAD - centered, mul);\n return below < WAD ? WAD - below : 0;\n }\n\n function getName() external pure override returns (string memory) { return \"no_cubic\"; }\n}\n", "sha256": "6dd68f77a70900319460215374143dbd5d924198450a46d9f5f91f6b1e3d35c3", "notes": { "searched": [ "Web search for AMM challenge / AMM fee challenge / ammchallenge.com / Optimization Arena combined with writeup, write-up, blog, strategy, fees, podcast, talk and interview.", "Targeted current-leader names 0xMai and Ponxter; historical names houseofjiao, adrianleb / Adrian le Bas, josusanmartin / Josu San Martin and unhedged21.", "Searched indexed X posts and mirrors, LinkedIn, Medium, Substack, Reddit, Farcaster/warpcast and YouTube. Followed the YQ reference from dipankar’s research post-mortem." ], "not_found": [ "No verified first-hand mechanism write-up by current leaders 0xMai or Ponxter, and no formula/code attributable to their current top scores.", "No entrant-specific talk, podcast, Reddit or Farcaster disclosure with an independently retrievable fee mechanism. Brady’s article links his own organizer talk/slides: https://www.benedict.dev/optimization-arena-learnings/slides ; no entrant talk transcript was recovered.", "Algebra provides qualitative mechanisms but not a reproducible policy; no code found in its article.", "Bruno I. profile surfaced in search indexing (https://hr.linkedin.com/in/bruno-i-8575b4b1), but direct fetch returned HTTP 999/authwall. Its mechanism and score details are withheld from verified sources/mechanisms, rather than presenting indexed snippets as freshly retrieved evidence.", "X search surfaced score-only announcements and speculative suggestions via mirrors. No adopted entrant fee formula was verified from a primary X post; none is promoted to a verified mechanism here.", "HackMD https://hackmd.io/6i1LNsRGRJ27YqTHjfO04A is an unattributed theory note, not an established entrant result. It is excluded from the entrant mechanisms and no policy is derived from it." ], "scope": "No strategy was designed, retuned or submitted. Source claims are attributed, not endorsed. Historical leaderboard claims are not current rankings. CH is the task-specified calibration reference, not a newly discovered write-up.", "setup": "Cloned the pinned simulator under test/scratch/amm-challenge; ran the supplied setup once. Wheel checksum passed; StarterStrategy validated. Executed none of the entrants’ scripts.", "method": "Independent arms: run_batch(candidate,VanillaStrategy,configs,8), then RS2 independently against VanillaStrategy on identical seeds. Config draws follow MatchRunner._build_configs; price seed=s, retail seed=s+1. 10000 steps. solc 0.8.24. Source bytes unchanged. Each source passes amm-match validate. Comparison is against RS2, not STACK; no claim of improvement over STACK is made.", "seeds": { "screen": "900000..900031 inclusive", "confirmation": "901000..901255 inclusive", "calibration": "CH only, 0..999 inclusive", "no_seeds_at_or_above_5000000": true }, "selection": "Exactly three strategy sources were collected for scoring: SigmaAdaptive from its post-mortem, YQ from that post-mortem’s cited entrant write-up, and the task-specified CH reference. All three proceed to 256 seeds; no fourth candidate or parameter tuning. Best means largest paired mean, even if all lose.", "decomposition": "Retail and arb deltas are retrospective reconstruction from stored per-step fair prices, previous spot and fees using pinned closed-form arb profit, as in the supplied harness. Retail=total edge-arb. Initial fees: SigmaAdaptive/YQ 3 bps, CH/RS2 4 bps. These are analysis inputs only; no strategy observes fair prices.", "gas": "Counters come from amm_sim_rs.gas_stats() after reset for each 16-seed chunk, for each independent arm. max_ok is the maximum Submission afterSwap successful gas across all chunks; oog sums Submission afterInitialize/afterSwap OOG. Extra counter fields record other halts. Any candidate OOG invalidates it.", "rs2_gas256": { "max_ok": 202386, "oog": 0, "init_max": 243099, "reverts": 0, "other_halts": 0, "tx_errors": 0, "bad_returns": 0 }, "rs2_gas32": { "max_ok": 197184, "oog": 0, "init_max": 243099, "reverts": 0, "other_halts": 0, "tx_errors": 0, "bad_returns": 0 }, "rs2_mean_edge256": 533.7351592392916, "CH_calibration": { "n": 1000, "mean_edge": 525.5258821780834, "sha256": "6dd68f77a70900319460215374143dbd5d924198450a46d9f5f91f6b1e3d35c3" }, "decomposition_reconciliation_max_abs": 4.218847493575595e-15 }, "code_url": "https://raw.githubusercontent.com/ChengHua926/amm-challenge/d34a5199c59a4c44d5336a25f1436b4ae1f21104/contracts/src/Strategy.sol" }