{"id":"https://openalex.org/W7160901185","doi":"https://doi.org/10.48550/arxiv.2605.10195","title":"Breaking the Reward Barrier: Accelerating Tree-of-Thought Reasoning via Speculative Exploration","display_name":"Breaking the Reward Barrier: Accelerating Tree-of-Thought Reasoning via Speculative Exploration","publication_year":2026,"publication_date":"2026-05-11","ids":{"openalex":"https://openalex.org/W7160901185","doi":"https://doi.org/10.48550/arxiv.2605.10195"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.10195","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10195","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.10195","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5081222727","display_name":"Shuzhang Zhong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhong, Shuzhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135915460","display_name":"Haochen Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Haochen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101328656","display_name":"Shengxuan Qiu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qiu, Shengxuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034442315","display_name":"Pengfei Zuo","orcid":"https://orcid.org/0000-0001-9982-5130"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zuo, Pengfei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135998597","display_name":"Runsheng Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Runsheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135927938","display_name":"Meng Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Meng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"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/T14347","display_name":"Big Data and Digital Economy","score":0.14740000665187836,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T14347","display_name":"Big Data and Digital Economy","score":0.14740000665187836,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T12292","display_name":"Graph Theory and Algorithms","score":0.10599999874830246,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11948","display_name":"Machine Learning in Materials Science","score":0.07599999755620956,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.7192000150680542},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.6226000189781189},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4894999861717224},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.4871000051498413},{"id":"https://openalex.org/keywords/synchronization","display_name":"Synchronization (alternating current)","score":0.45980000495910645},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.4018000066280365},{"id":"https://openalex.org/keywords/parallelism","display_name":"Parallelism (grammar)","score":0.3806000053882599},{"id":"https://openalex.org/keywords/resource","display_name":"Resource (disambiguation)","score":0.35409998893737793},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.3418000042438507}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7506999969482422},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.7192000150680542},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6226000189781189},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4894999861717224},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.4871000051498413},{"id":"https://openalex.org/C2778562939","wikidata":"https://www.wikidata.org/wiki/Q1298791","display_name":"Synchronization (alternating current)","level":3,"score":0.45980000495910645},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.4018000066280365},{"id":"https://openalex.org/C2781172179","wikidata":"https://www.wikidata.org/wiki/Q853109","display_name":"Parallelism (grammar)","level":2,"score":0.3806000053882599},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.35409998893737793},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3418000042438507},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.33550000190734863},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.33090001344680786},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3278999924659729},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.3257000148296356},{"id":"https://openalex.org/C159032336","wikidata":"https://www.wikidata.org/wiki/Q2488768","display_name":"Non-monotonic logic","level":2,"score":0.3012000024318695},{"id":"https://openalex.org/C125583679","wikidata":"https://www.wikidata.org/wiki/Q755673","display_name":"Search algorithm","level":2,"score":0.2824000120162964},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.2816999852657318},{"id":"https://openalex.org/C2780310539","wikidata":"https://www.wikidata.org/wiki/Q12547192","display_name":"Imperfect","level":2,"score":0.2750000059604645},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.27390000224113464},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.27309998869895935},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.26649999618530273},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.2637999951839447},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.2624000012874603},{"id":"https://openalex.org/C311688","wikidata":"https://www.wikidata.org/wiki/Q2393193","display_name":"Time complexity","level":2,"score":0.2565999925136566},{"id":"https://openalex.org/C19768560","wikidata":"https://www.wikidata.org/wiki/Q320727","display_name":"Dependency (UML)","level":2,"score":0.2563999891281128},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25189998745918274}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.10195","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10195","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.10195","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10195","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Tree-of-Thought":[0],"(ToT)":[1],"reasoning":[2,73,79,176],"structures":[3],"Large":[4],"Language":[5],"Model":[6],"(LLM)":[7],"inference":[8],"as":[9],"a":[10,34,140,204],"tree-based":[11],"search,":[12],"demonstrating":[13],"strong":[14],"potential":[15],"for":[16,55,139,173,217,221],"solving":[17],"complex":[18],"mathematical":[19],"and":[20,46,98,110,128,136,153,160,209],"programming":[21],"tasks.":[22],"However,":[23],"its":[24],"efficiency":[25,66],"is":[26],"constrained":[27],"by":[28,38],"the":[29,65,78,87,150,196,214],"reward":[30,88],"dependency":[31],"barrier":[32],"--":[33],"synchronization":[35,89],"bottleneck":[36],"caused":[37],"sequential":[39],"reward-guided":[40],"exploration":[41],"that":[42,77,165],"limits":[43],"search":[44,142],"parallelism":[45,215],"introduces":[47],"substantial":[48],"latency.":[49],"Prior":[50],"system":[51],"optimizations,":[52],"mainly":[53],"designed":[54],"linear":[56],"Chain-of-Thought":[57],"(CoT)":[58],"reasoning,":[59,212],"cannot":[60],"address":[61],"these":[62],"challenges,":[63],"leaving":[64],"of":[67,114,149,188,198],"ToT":[68,72,158,175,211],"underexplored.":[69],"To":[70],"enhance":[71],"efficiency,":[74],"we":[75,95],"observe":[76],"paths":[80],"can":[81],"be":[82],"explored":[83],"speculatively":[84],"to":[85,108,120,133,190],"break":[86],"barrier.":[90],"Therefore,":[91],"in":[92],"this":[93],"paper,":[94],"propose":[96],"SPEX":[97,146,166,179,202],"introduce":[99],"three":[100],"key":[101],"techniques:":[102],"(i)":[103],"intra-query":[104],"speculative":[105,122,183],"path":[106],"selection":[107],"predict":[109],"expand":[111],"high-potential":[112],"branches":[113,138],"ToT,":[115],"(ii)":[116],"inter-query":[117],"budget":[118],"allocation":[119,124],"balance":[121],"resource":[123],"across":[125,156],"queries":[126],"dynamically,":[127],"(iii)":[129],"adaptive":[130],"early":[131],"termination":[132],"prune":[134],"deep":[135],"redundant":[137],"skewed":[141],"tree.":[143],"We":[144],"implement":[145],"on":[147],"top":[148],"SGLang":[151],"framework":[152],"evaluate":[154],"it":[155],"diverse":[157],"algorithms":[159],"LLMs.":[161,222],"Extensive":[162],"experiments":[163],"show":[164],"achieves":[167],"$1.2":[168],"\\sim":[169],"3":[170],"\\times$":[171],"speedup":[172],"different":[174],"algorithms.":[177],"Moreover,":[178],"synergizes":[180],"with":[181],"token-level":[182],"decoding,":[184],"achieving":[185],"cumulative":[186],"speedups":[187],"up":[189],"$4.1\\times$.":[191],"Ablation":[192],"studies":[193],"further":[194],"confirm":[195],"contributions":[197],"each":[199],"technique.":[200],"Overall,":[201],"represents":[203],"significant":[205],"step":[206],"toward":[207],"efficient":[208],"scalable":[210],"unlocking":[213],"required":[216],"high-performance":[218],"inference-time":[219],"scaling":[220]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-13T00:00:00"}