{"id":"https://openalex.org/W7171258508","doi":"https://doi.org/10.48550/arxiv.2609.28717","title":"HeteroReason: Heterogeneous FPGA-GPU Acceleration for Disaggregated Speculative Reasoning","display_name":"HeteroReason: Heterogeneous FPGA-GPU Acceleration for Disaggregated Speculative Reasoning","publication_year":2026,"publication_date":"2026-07-25","ids":{"openalex":"https://openalex.org/W7171258508","doi":"https://doi.org/10.48550/arxiv.2609.28717"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.28717","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.28717","pdf_url":"https://arxiv.org/pdf/2609.28717","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2609.28717","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5115600242","display_name":"Zehuan Zhang","orcid":"https://orcid.org/0009-0006-0607-060X","observed_orcids":["https://orcid.org/0009-0006-0607-060X"]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Zhang, Zehuan","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Deng, Quan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]},{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN","GB"],"is_corresponding":false,"raw_author_name":"Deng, Quan","raw_affiliation_strings":["Imperial College London","Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]},{"raw_affiliation_string":"Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Ren, Zibo","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Ren, Zibo","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Chen, Hao Mark","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Chen, Hao Mark","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Li, Guoyu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Li, Guoyu","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Hu, Xuchun","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Hu, Xuchun","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Coutinho, Jose G. F.","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Coutinho, Jose G. F.","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Guo, Ce","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Guo, Ce","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Luk, Wayne","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Luk, Wayne","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Que, Zhiqiang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I36234482","display_name":"University of Bristol","ror":"https://ror.org/0524sp257","country_code":"GB","type":"education","lineage":["https://openalex.org/I36234482"]},{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Que, Zhiqiang","raw_affiliation_strings":["Imperial College London","University of Bristol"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]},{"raw_affiliation_string":"University of Bristol","institution_ids":["https://openalex.org/I36234482"]}]},{"author_position":"last","author":{"id":null,"display_name":"Fan, Hongxiang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Fan, Hongxiang","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T10904","display_name":"Embedded Systems Design Techniques","score":0.76419997215271,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10904","display_name":"Embedded Systems Design Techniques","score":0.76419997215271,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.17419999837875366,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.012500000186264515,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/speculative-reasoning","display_name":"speculative reasoning","score":0.9179999828338623},{"id":"https://openalex.org/keywords/large-reasoning-models","display_name":"large reasoning models","score":0.8259999752044678},{"id":"https://openalex.org/keywords/chain-of-thought","display_name":"chain-of-thought","score":0.8769999742507935},{"id":"https://openalex.org/keywords/algorithm-hardware-co-design","display_name":"algorithm-hardware co-design","score":0.8349999785423279},{"id":"https://openalex.org/keywords/process-reward-models","display_name":"process reward models","score":0.753000020980835},{"id":"https://openalex.org/keywords/backtracking","display_name":"backtracking","score":0.621999979019165},{"id":"https://openalex.org/keywords/error-propagation","display_name":"error propagation","score":0.5019999742507935},{"id":"https://openalex.org/keywords/energy-efficiency","display_name":"energy efficiency","score":0.7760000228881836}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8306999802589417},{"id":"https://openalex.org/C89288958","wikidata":"https://www.wikidata.org/wiki/Q7301504","display_name":"Reasoning system","level":2,"score":0.5745999813079834},{"id":"https://openalex.org/C86827895","wikidata":"https://www.wikidata.org/wiki/Q7098582","display_name":"Opportunistic reasoning","level":4,"score":0.5698000192642212},{"id":"https://openalex.org/C159032336","wikidata":"https://www.wikidata.org/wiki/Q2488768","display_name":"Non-monotonic logic","level":2,"score":0.5530999898910522},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5238999724388123},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.5153999924659729},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.5011000037193298},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.4943000078201294},{"id":"https://openalex.org/C2778562939","wikidata":"https://www.wikidata.org/wiki/Q1298791","display_name":"Synchronization (alternating current)","level":3,"score":0.4756999909877777},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.4410000145435333},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.42250001430511475},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.3921000063419342},{"id":"https://openalex.org/C156884757","wikidata":"https://www.wikidata.org/wiki/Q798554","display_name":"Backtracking","level":2,"score":0.3901999890804291},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.3659999966621399},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.34369999170303345},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.3402000069618225},{"id":"https://openalex.org/C107848011","wikidata":"https://www.wikidata.org/wiki/Q4680756","display_name":"Adaptive reasoning","level":4,"score":0.3400999903678894},{"id":"https://openalex.org/C115086926","wikidata":"https://www.wikidata.org/wiki/Q17004651","display_name":"Causal reasoning","level":3,"score":0.3384000062942505},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.33180001378059387},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3230000138282776},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.3215999901294708},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.32019999623298645},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.3050999939441681},{"id":"https://openalex.org/C97364631","wikidata":"https://www.wikidata.org/wiki/Q484284","display_name":"Deductive reasoning","level":2,"score":0.28690001368522644},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.27469998598098755},{"id":"https://openalex.org/C9616225","wikidata":"https://www.wikidata.org/wiki/Q3929429","display_name":"Semantic reasoner","level":2,"score":0.2702000141143799},{"id":"https://openalex.org/C103057564","wikidata":"https://www.wikidata.org/wiki/Q4751139","display_name":"Analytic reasoning","level":3,"score":0.2624000012874603},{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C183521366","wikidata":"https://www.wikidata.org/wiki/Q7256422","display_name":"Psychology of reasoning","level":4,"score":0.2572000026702881},{"id":"https://openalex.org/C111498074","wikidata":"https://www.wikidata.org/wiki/Q173326","display_name":"Formal verification","level":2,"score":0.25600001215934753},{"id":"https://openalex.org/C129916263","wikidata":"https://www.wikidata.org/wiki/Q1141183","display_name":"Backward chaining","level":4,"score":0.25290000438690186},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.25279998779296875},{"id":"https://openalex.org/C2780385302","wikidata":"https://www.wikidata.org/wiki/Q367158","display_name":"Protocol (science)","level":3,"score":0.25209999084472656}],"mesh":[],"locations_count":3,"locations":[{"id":"pmh:oai:arXiv.org:2609.28717","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.28717","pdf_url":"https://arxiv.org/pdf/2609.28717","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.28717","doi":"https://doi.org/10.48550/arxiv.2609.28717","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.28717","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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"},{"id":"doi:10.5281/zenodo.21548289","doi":"https://doi.org/10.5281/zenodo.21548289","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.21548289","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":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.28717","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.28717","pdf_url":"https://arxiv.org/pdf/2609.28717","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G3496335909","display_name":"Centre for Spatial Computational Learning","funder_award_id":"EP/S030069/1","funder_id":"https://openalex.org/F4320334627","funder_display_name":"Engineering and Physical Sciences Research Council"},{"id":"https://openalex.org/G488418063","display_name":"SONNETS: Scalability Oriented Novel Network of Event Triggered Systems","funder_award_id":"EP/X036006/1","funder_id":"https://openalex.org/F4320334627","funder_display_name":"Engineering and Physical Sciences Research Council"},{"id":"https://openalex.org/G514660964","display_name":"DART: Design Accelerators by Regulating Transformations","funder_award_id":"EP/V028251/1","funder_id":"https://openalex.org/F4320334627","funder_display_name":"Engineering and Physical Sciences Research Council"}],"funders":[{"id":"https://openalex.org/F4320314731","display_name":"UK Research and Innovation","ror":"https://ror.org/001aqnf71"},{"id":"https://openalex.org/F4320322064","display_name":"Korea Institute for Advancement of Technology","ror":"https://ror.org/015w1qa96"},{"id":"https://openalex.org/F4320334627","display_name":"Engineering and Physical Sciences Research Council","ror":"https://ror.org/0439y7842"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7171258508.pdf","grobid_xml":"https://content.openalex.org/works/W7171258508.grobid-xml"},"referenced_works_count":44,"referenced_works":["https://openalex.org/W3134642945","https://openalex.org/W3210904070","https://openalex.org/W4391124766","https://openalex.org/W4392678661","https://openalex.org/W4392822465","https://openalex.org/W4396758996","https://openalex.org/W4399317989","https://openalex.org/W4399454054","https://openalex.org/W4401306886","https://openalex.org/W4401307635","https://openalex.org/W4402670720","https://openalex.org/W4403746689","https://openalex.org/W4405766390","https://openalex.org/W4406735009","https://openalex.org/W4407091778","https://openalex.org/W4414581587","https://openalex.org/W4414684785","https://openalex.org/W4415070954","https://openalex.org/W4415108962","https://openalex.org/W4415328803","https://openalex.org/W4416035986","https://openalex.org/W4416046236","https://openalex.org/W4416429558","https://openalex.org/W4416623251","https://openalex.org/W4416649973","https://openalex.org/W4416892671","https://openalex.org/W4416955037","https://openalex.org/W4417087151","https://openalex.org/W4417088883","https://openalex.org/W4417299515","https://openalex.org/W4417299931","https://openalex.org/W6891807915","https://openalex.org/W6910522192","https://openalex.org/W6948351282","https://openalex.org/W7119698293","https://openalex.org/W7124448701","https://openalex.org/W7125522847","https://openalex.org/W7127639249","https://openalex.org/W7134839917","https://openalex.org/W7162638753","https://openalex.org/W7166833673","https://openalex.org/W7168260600","https://openalex.org/W7196954301","https://openalex.org/W7214236243"],"related_works":["https://openalex.org/W4417299515","https://openalex.org/W7162638753","https://openalex.org/W4417088883","https://openalex.org/W4416033863","https://openalex.org/W4417087151","https://openalex.org/W7131319044","https://openalex.org/W4417126394","https://openalex.org/W4400104534","https://openalex.org/W4414684785","https://openalex.org/W4416892671"],"abstract_inverted_index":{"Large":[0],"Reasoning":[1],"Models":[2],"(LRMs)":[3],"have":[4,24],"achieved":[5],"state-of-the-art":[6],"performance":[7],"in":[8,27],"reasoning":[9,22,62,68,82,148,152,234,245],"tasks":[10],"by":[11,41,236],"utilizing":[12],"Chain-of-Thought":[13],"(CoT)":[14],"reasoning.":[15,126],"To":[16,108,201],"achieve":[17,181],"fast":[18],"execution":[19,221],"speed,":[20],"speculative":[21,61,125],"techniques":[23],"been":[25],"introduced":[26],"prior":[28],"work,":[29],"which":[30,70,184],"adopt":[31],"a":[32,49,65,133,208,219,224],"lightweight":[33],"draft":[34,159],"model":[35,52,160],"for":[36,46,53,123],"candidate":[37],"token":[38,194],"generation":[39],"followed":[40],"process":[42],"reward":[43],"models":[44,172],"(PRMs)":[45],"verification":[47,97],"and":[48,73,95,145,170,211,243],"strong":[50],"target":[51,171],"refinements.":[54],"This":[55],"paper":[56],"identifies":[57],"that":[58,136],"the":[59,128,138,155,158,164,168,216,230],"existing":[60],"paradigm":[63,120],"follows":[64],"strictly":[66],"forward-only":[67],"trajectory,":[69],"lacks":[71],"robustness":[72],"can":[74,102],"lead":[75,103],"to":[76,104,140,163,180,188,196,223,238],"severe":[77,105],"error":[78],"propagation":[79],"if":[80],"early":[81],"steps":[83],"are":[84],"suboptimal.":[85],"Furthermore,":[86],"executing":[87],"these":[88],"disparate":[89],"inference":[90,119],"schemes,":[91],"including":[92],"sequential":[93,204,220],"drafting":[94],"parallel":[96],"on":[98,173],"homogeneous":[99],"GPU":[100],"platforms,":[101],"resource":[106],"underutilization.":[107],"address":[109],"this,":[110],"we":[111,131,206],"propose":[112,207],"HeteroReason,":[113],"an":[114],"algorithm-hardware":[115],"co-designed":[116],"heterogeneous":[117],"FPGA-GPU":[118],"specifically":[121],"tailored":[122],"LRM":[124],"At":[127,154],"algorithmic":[129],"level,":[130,157],"introduce":[132],"backtracking-enhanced":[134],"workflow":[135,177],"enables":[137],"system":[139,156,217],"recover":[141],"from":[142,218],"low-quality":[143],"states":[144],"explore":[146],"alternative":[147],"trajectories,":[149],"significantly":[150],"improving":[151],"robustness.":[153],"is":[161,178],"offloaded":[162],"FPGA":[165],"while":[166],"deploying":[167],"PRM":[169],"GPUs.":[174],"A":[175],"specialized":[176],"optimized":[179],"prefill-decode":[182],"disaggregation,":[183],"exploits":[185],"shadow":[186],"synchronization":[187,199],"overlap":[189],"GPU-side":[190],"refinements":[191],"with":[192],"FPGA-side":[193],"updates":[195],"effectively":[197],"hide":[198],"latency.":[200],"mitigate":[202],"inherent":[203],"constraints,":[205],"step-ahead":[209],"speculation":[210],"refinement":[212],"scheduling":[213],"scheme,":[214],"transitioning":[215],"scheme":[222],"parallelized":[225],"pipeline.":[226],"Experimental":[227],"evaluations":[228],"demonstrate":[229],"backtracking":[231],"mechanism":[232],"improves":[233],"accuracy":[235],"up":[237],"$4.4\\%$":[239],"across":[240],"diverse":[241],"mathematical":[242],"logical":[244],"benchmarks.":[246]},"counts_by_year":[],"updated_date":"2026-11-08T12:12:28.789561","created_date":"2026-07-26T00:00:00"}