{"id":85,"notebook":86,"tag":"#086","status":"published","suite":"defi-maths","suiteTitle":"DeFi maths","exactMeans":"the final number is right to the cent (to the whole token for bonding-curve buys)","base":{"model":"claude-haiku-5-5","thinking":"off"},"pairsDone":432,"maxPairs":432,"beaconRound":6625283,"beaconAt":"2026-11-09T01:58:30.000Z","beacon":{"chain":"8990e7a9aaed2ffed73dbd7092123d6f289930540d7651336225dc172e51b2ce","round":6625283,"signature":"b644ddf4272d6ddfa29a498f6216c75a8102916cbfaaeb392afa4d283739fb161a28d4967e09f30d3b420102e92c79de14330721a8be02371a8da2924f40cb8e9d39e7b41714890869d39b46bf8d217814441d64bb944808f9f5268d7aefcf9b","randomness":"f2e228ad48c2d7e98a128ba35d1d847b1270e901cf4e9172ef5da5605472593a","publishedAt":"2026-11-09T01:58:30.000Z"},"seed":"drand:6625283:f2e228ad48c2d7e98a128ba35d1d847b1270e901cf4e9172ef5da5605472593a","preregHash":"f9eb5553a0186129a186403e7e58da6d7562ba8a94b42320a6ba37e07e47cad8","prereg":{"lab":"OPSIN","arms":[{"role":"control-a","fingerprint":"5fa877820c8d2c9fa461a84538aa019d7052d1f9e93ca9600c5f138288b0a3a9"},{"role":"control-b","fingerprint":"5fa877820c8d2c9fa461a84538aa019d7052d1f9e93ca9600c5f138288b0a3a9"},{"part":{"kind":"line","text":"The weather outside is mild today."},"role":"placebo","fingerprint":"8be85d978fc89a9dbbc2d69472edc569246b659bdc5f2b5a0c2c8fd66df434c7"},{"part":{"kind":"line","text":"Write out each calculation step on its own line with the numbers substituted in before giving the final answer."},"role":"part","label":"Show every step","proposal":358,"fingerprint":"b14374fcb113efa1e684083852207878cf85868a7afac93e98d43ae868929a2e"},{"part":{"kind":"line","text":"Carry full precision through every intermediate result and round only the final answer to the requested format."},"role":"part","label":"Keep full precision","proposal":359,"fingerprint":"320e56ae5c2967c726b3861112c5183b47671df51b57ced00d117180a6c16ce2"},{"part":{"kind":"line","text":"After computing the result, redo only the last arithmetic operation from the written inputs and correct it if it differs."},"role":"part","label":"Recompute final step","proposal":360,"fingerprint":"a4909dc9354a134c37f172ea6608f0496f7eb44f825329e280ef5591ef867936"}],"base":{"model":"claude-haiku-5-5","thinking":"off"},"kind":"cycle","rule":{"cost":"each part's cost over the controls' mean, both at list prices: a batched run was paid half, so it is doubled","INERT":"adjusted interval entirely inside ±5 points","errors":"a task with an errored run in any compared arm is left out of that comparison","metric":"exact: the final number is right to the cent (to the whole token for bonding-curve buys)","EXCITES":"adjusted interval entirely above 0","control":"each part against the per-task mean of two unchanged runs of the base agent","interim":"before the last look, a part stops early only if |z| ≥ 3.29; INERT is decided at the last look only; a part that stops early with EXCITES or INHIBITS is provisional until a later notebook confirms it at full size","placebo":"a neutral line chosen by the beacon round's number from a fixed public list","version":2,"INHIBITS":"adjusted interval entirely below 0","interval":"percentile bootstrap over tasks, 20,000 resamples, seeded from the beacon seed and the proposal id, at 1 − 0.05/4: 4 arms (the parts and the placebo) are compared with the controls, so the chance that any arm of this notebook flags by luck stays at or under 5%","UNRESOLVED":"anything else","confirmations":"an arm that re-measures an earlier early stop never stops early","versusPlacebo":"each part is also compared with the placebo on the same tasks, 95% interval, reported beside its verdict"},"looks":[48,96,192,432],"suite":{"id":"defi-maths","version":"tasks-2","exactMeans":"the final number is right to the cent (to the whole token for bonding-curve buys)"},"beacon":{"rule":"first drand round at least 600 seconds after this preregistration","chain":"8990e7a9aaed2ffed73dbd7092123d6f289930540d7651336225dc172e51b2ce","round":6625283},"prices":"2026-11-04","version":2,"maxPairs":432,"foresight":{"arms":[{"basis":"new","rates":{"INERT":0.1292,"EXCITES":0.2568,"INHIBITS":0.1695,"UNRESOLVED":0.4445},"record":{"INERT":0.0146,"EXCITES":0.282,"INHIBITS":0.1758,"UNRESOLVED":0.5276},"chances":{"INERT":0.0794,"EXCITES":0.2462,"INHIBITS":0.1781,"UNRESOLVED":0.4963},"measured":0,"proposal":358},{"basis":"new","rates":{"INERT":0.1292,"EXCITES":0.2568,"INHIBITS":0.1695,"UNRESOLVED":0.4445},"record":{"INERT":0.0169,"EXCITES":0.284,"INHIBITS":0.171,"UNRESOLVED":0.5281},"chances":{"INERT":0.0808,"EXCITES":0.2474,"INHIBITS":0.1752,"UNRESOLVED":0.4966},"measured":0,"proposal":359},{"basis":"new","rates":{"INERT":0.1292,"EXCITES":0.2568,"INHIBITS":0.1695,"UNRESOLVED":0.4445},"record":{"INERT":0.0189,"EXCITES":0.26,"INHIBITS":0.162,"UNRESOLVED":0.5591},"chances":{"INERT":0.0761,"EXCITES":0.2272,"INHIBITS":0.164,"UNRESOLVED":0.5327},"measured":0,"proposal":360}],"version":2,"weights":{"new":[0.35,0.6,0.05],"stack":[0.33,0.33,0.33],"measured":[0,1,0]}},"predictor":2,"writtenAt":"2026-11-09T01:48:14.410Z","predictedBy":"claude-sonnet-5-5","predictions":[{"p":0.45,"why":"Thinking off, so stepwise helps, but base likely already shows steps","effect":3,"verdict":"UNRESOLVED","proposal":358},{"p":0.45,"why":"Precision helps rounding errors in DeFi maths; modest, uncertain gain","effect":3,"verdict":"UNRESOLVED","proposal":359},{"p":0.5,"why":"Checking only last step catches little; small effect, wide interval","effect":1,"verdict":"UNRESOLVED","proposal":360}]},"createdAt":"2026-11-09T01:48:15.426Z","publishedAt":"2026-11-09T03:11:14.776Z","error":null,"controlsTripped":false,"baseExact":0.6146,"arms":[{"id":616,"role":"control-a","label":"Control A","kind":"control-a","text":null,"proposalId":null,"origin":null,"stoppedAt":null,"done":432,"exact":265,"prediction":null,"foresight":null,"verdict":null},{"id":617,"role":"control-b","label":"Control B","kind":"control-b","text":null,"proposalId":null,"origin":null,"stoppedAt":null,"done":432,"exact":266,"prediction":null,"foresight":null,"verdict":{"id":524,"verdict":"INERT","effect":0.23,"lo":-4.17,"hi":4.4,"n":432,"costRatio":null,"stoppedEarly":false,"safety":null,"rule":2,"level":null,"raw95":null,"vsPlacebo":null,"provisional":false,"confirms":null,"stack":null}},{"id":618,"role":"placebo","label":"Placebo","kind":"placebo","text":"The weather outside is mild today.","proposalId":null,"origin":null,"stoppedAt":null,"done":432,"exact":258,"prediction":null,"foresight":null,"verdict":{"id":525,"verdict":"UNRESOLVED","effect":-1.74,"lo":-6.71,"hi":3.36,"n":432,"costRatio":0.949,"stoppedEarly":false,"safety":null,"rule":2,"level":0.9875,"raw95":{"hi":2.2,"lo":-5.67,"verdict":"UNRESOLVED"},"vsPlacebo":null,"provisional":false,"confirms":null,"stack":null}},{"id":619,"role":"part","label":"Show every step","kind":"line","text":"Write out each calculation step on its own line with the numbers substituted in before giving the final answer.","proposalId":358,"origin":"OPSIN's pick","stoppedAt":null,"done":432,"exact":237,"prediction":{"verdict":"UNRESOLVED","p":0.45,"why":"Thinking off, so stepwise helps, but base likely already shows steps"},"foresight":{"basis":"new","measured":0,"chances":{"INERT":0.0794,"EXCITES":0.2462,"INHIBITS":0.1781,"UNRESOLVED":0.4963}},"verdict":{"id":526,"verdict":"INHIBITS","effect":-6.6,"lo":-11.69,"hi":-1.39,"n":432,"costRatio":1.098,"stoppedEarly":false,"safety":null,"rule":2,"level":0.9875,"raw95":{"hi":-2.43,"lo":-10.53,"verdict":"INHIBITS"},"vsPlacebo":{"n":432,"hi":-0.46,"lo":-9.49,"effect":-4.86},"provisional":false,"confirms":null,"stack":null}},{"id":620,"role":"part","label":"Keep full precision","kind":"line","text":"Carry full precision through every intermediate result and round only the final answer to the requested format.","proposalId":359,"origin":"OPSIN's pick","stoppedAt":null,"done":432,"exact":281,"prediction":{"verdict":"UNRESOLVED","p":0.45,"why":"Precision helps rounding errors in DeFi maths; modest, uncertain gain"},"foresight":{"basis":"new","measured":0,"chances":{"INERT":0.0808,"EXCITES":0.2474,"INHIBITS":0.1752,"UNRESOLVED":0.4966}},"verdict":{"id":527,"verdict":"UNRESOLVED","effect":3.59,"lo":-1.39,"hi":8.68,"n":432,"costRatio":1.137,"stoppedEarly":false,"safety":null,"rule":2,"level":0.9875,"raw95":{"hi":7.52,"lo":-0.35,"verdict":"UNRESOLVED"},"vsPlacebo":{"n":432,"hi":9.95,"lo":0.69,"effect":5.32},"provisional":false,"confirms":null,"stack":null}},{"id":621,"role":"part","label":"Recompute final step","kind":"line","text":"After computing the result, redo only the last arithmetic operation from the written inputs and correct it if it differs.","proposalId":360,"origin":"OPSIN's pick","stoppedAt":null,"done":432,"exact":291,"prediction":{"verdict":"UNRESOLVED","p":0.5,"why":"Checking only last step catches little; small effect, wide interval"},"foresight":{"basis":"new","measured":0,"chances":{"INERT":0.0761,"EXCITES":0.2272,"INHIBITS":0.164,"UNRESOLVED":0.5327}},"verdict":{"id":528,"verdict":"EXCITES","effect":5.9,"lo":1.5,"hi":10.42,"n":432,"costRatio":1.268,"stoppedEarly":false,"safety":null,"rule":2,"level":0.9875,"raw95":{"hi":9.38,"lo":2.31,"verdict":"EXCITES"},"vsPlacebo":{"n":432,"hi":11.57,"lo":3.7,"effect":7.64},"provisional":false,"confirms":null,"stack":null}}]}