{"id":"https://openalex.org/W7166320120","doi":"https://doi.org/10.5281/zenodo.20970204","title":"The Broadcast Ceiling: Information and Reliability Limits of Advantage Estimators in Long-Horizon RL","display_name":"The Broadcast Ceiling: Information and Reliability Limits of Advantage Estimators in Long-Horizon RL","publication_year":2026,"publication_date":"2026-06-27","ids":{"openalex":"https://openalex.org/W7166320120","doi":"https://doi.org/10.5281/zenodo.20970204"},"language":"en","primary_location":{"id":"pmh:doi:10.5281/zenodo.20970204","doi":null,"is_oa":true,"landing_page_url":"https://github.com/Limes-Labs/the-broadcast-ceiling/releases/tag/v0.1.0","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"volume":null,"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://github.com/Limes-Labs/the-broadcast-ceiling/releases/tag/v0.1.0","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139487183","display_name":"Francesco Giannicola","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Giannicola, Francesco","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5139487183"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9926999807357788,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9926999807357788,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.0007999999797903001,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.000699999975040555,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/advantage-estimation","display_name":"advantage estimation","score":0.6620000004768372},{"id":"https://openalex.org/keywords/credit-assignment","display_name":"credit assignment","score":0.6679999828338623},{"id":"https://openalex.org/keywords/estimator-reliability","display_name":"estimator reliability","score":0.460999995470047},{"id":"https://openalex.org/keywords/information-theoretic-limits","display_name":"information-theoretic limits","score":0.5860000252723694},{"id":"https://openalex.org/keywords/policy-gradient","display_name":"policy gradient","score":0.5540000200271606}],"concepts":[{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.776199996471405},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6237999796867371},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.5557000041007996},{"id":"https://openalex.org/C36299963","wikidata":"https://www.wikidata.org/wiki/Q1369844","display_name":"Observability","level":2,"score":0.5390999913215637},{"id":"https://openalex.org/C55166926","wikidata":"https://www.wikidata.org/wiki/Q2892946","display_name":"Oracle","level":2,"score":0.42100000381469727},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4115000069141388},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.4083000123500824},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.4075999855995178},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.3871999979019165},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.36649999022483826},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.352400004863739},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.3490999937057495},{"id":"https://openalex.org/C2778565505","wikidata":"https://www.wikidata.org/wiki/Q2207566","display_name":"Spec#","level":2,"score":0.34700000286102295},{"id":"https://openalex.org/C2780695315","wikidata":"https://www.wikidata.org/wiki/Q3799040","display_name":"Unobservable","level":2,"score":0.3395000100135803},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.33500000834465027},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.3336000144481659},{"id":"https://openalex.org/C57691317","wikidata":"https://www.wikidata.org/wiki/Q1289248","display_name":"Scalar (mathematics)","level":2,"score":0.3260999917984009},{"id":"https://openalex.org/C34862557","wikidata":"https://www.wikidata.org/wiki/Q178985","display_name":"Ode","level":2,"score":0.31360000371932983},{"id":"https://openalex.org/C2779664074","wikidata":"https://www.wikidata.org/wiki/Q3518405","display_name":"Terminal (telecommunication)","level":2,"score":0.30570000410079956},{"id":"https://openalex.org/C199521495","wikidata":"https://www.wikidata.org/wiki/Q181487","display_name":"Audit","level":2,"score":0.27950000762939453},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.2777999937534332},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.27300000190734863},{"id":"https://openalex.org/C2781311116","wikidata":"https://www.wikidata.org/wiki/Q83306","display_name":"Group (periodic table)","level":2,"score":0.25540000200271606},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.25529998540878296},{"id":"https://openalex.org/C201729545","wikidata":"https://www.wikidata.org/wiki/Q2193887","display_name":"Reliability theory","level":3,"score":0.25519999861717224}],"mesh":[],"locations_count":4,"locations":[{"id":"pmh:doi:10.5281/zenodo.20970204","doi":null,"is_oa":true,"landing_page_url":"https://github.com/Limes-Labs/the-broadcast-ceiling/releases/tag/v0.1.0","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},{"id":"pmh:oai:zenodo.org:20970205","doi":null,"is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.20970205","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/preprint"},{"id":"doi:10.5281/zenodo.20970204","doi":"https://doi.org/10.5281/zenodo.20970204","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.20970204","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"},{"id":"doi:10.5281/zenodo.20970205","doi":"https://doi.org/10.5281/zenodo.20970205","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.20970205","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:doi:10.5281/zenodo.20970204","doi":null,"is_oa":true,"landing_page_url":"https://github.com/Limes-Labs/the-broadcast-ceiling/releases/tag/v0.1.0","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Long-horizon":[0],"language-model":[1],"RL":[2],"often":[3],"receives":[4],"terminal":[5],"trajectory":[6,20],"rewards":[7],"while":[8,136],"updating":[9],"token-":[10],"or":[11,134],"step-level":[12],"actions.":[13],"This":[14],"paper":[15],"studies":[16],"when":[17,25,119,129],"a":[18,40,44,182,194],"scalar":[19],"coefficient":[21,51],"is":[22,29,52,95,149,180,193],"enough":[23],"and":[24,39,67,73,77,86,121,167,186,201],"reliable":[26],"state-varying":[27],"credit":[28,110,162],"needed.":[30],"We":[31],"decompose":[32],"estimator":[33],"error":[34],"into":[35],"an":[36,188],"information":[37],"term":[38],"reliability":[41],"term,":[42],"proving":[43],"broadcast":[45],"ceiling":[46],"for":[47],"any":[48],"method":[49],"whose":[50],"token-invariant":[53],"under":[54],"the":[55,143],"available":[56],"group":[57,125],"context.":[58],"CPU-only":[59],"finite-MDP":[60],"audits":[61,200],"then":[62],"compare":[63],"group-relative":[64],"broadcasts,":[65],"prefix":[66],"structural":[68,137],"baselines,":[69],"sampled":[70],"values,":[71],"learned":[72],"oracle":[74],"value":[75,130],"TD,":[76],"policy-implied":[78],"actor":[79],"coefficients,":[80],"with":[81,198],"downstream":[82],"checks":[83],"including":[84],"tabular":[85],"tiny":[87],"autoregressive":[88],"sequence-policy":[89],"training.":[90],"The":[91,147],"central":[92],"empirical":[93],"finding":[94],"conditional:":[96],"trajectory-level":[97],"coefficients":[98],"can":[99,106,127,139],"give":[100],"high":[101],"policy-gradient":[102],"cosine,":[103],"but":[104],"they":[105],"be":[107,159],"poor":[108],"local":[109,145],"explanations":[111],"in":[112],"heterogeneous":[113],"traces;":[114],"value-style":[115],"TD":[116],"reduces":[117],"variance":[118],"coverage":[120],"observability":[122],"are":[123,132],"adequate;":[124],"methods":[126,138],"win":[128],"estimates":[131],"blind":[133],"unsupported,":[135],"recover":[140],"part":[141],"of":[142],"lost":[144],"information.":[146],"conclusion":[148],"not":[150,181,187],"that":[151],"PPO":[152],"universally":[153],"beats":[154],"GRPO.":[155],"Estimator":[156],"choice":[157],"should":[158],"made":[160],"from":[161],"heterogeneity,":[163],"observability,":[164],"support,":[165],"drift,":[166],"compute":[168],"cost.":[169],"Canonical":[170],"GitHub":[171],"release:":[172],"https://github.com/Limes-Labs/the-broadcast-ceiling/releases/tag/v0.1.0Repository:":[173],"https://github.com/Limes-Labs/the-broadcast-ceilingCanonical":[174],"PDF:":[175],"https://github.com/Limes-Labs/the-broadcast-ceiling/releases/download/v0.1.0/the-broadcast-ceiling-v0.1.0.pdf":[176],"Limitations:":[177],"this":[178],"release":[179],"transformer-scale":[183],"PPO-vs-GRPO":[184],"benchmark":[185],"LLM":[189],"leaderboard":[190],"result.":[191],"It":[192],"focused":[195],"mechanism/theory":[196],"artifact":[197],"synthetic":[199],"explicit":[202],"external-validity":[203],"limitations.":[204]},"counts_by_year":[],"updated_date":"2026-10-07T09:51:55.459714","created_date":"2026-06-28T00:00:00"}