{"id":"https://openalex.org/W7166846202","doi":"https://doi.org/10.18653/v1/2026.findings-acl.918","title":"On the Step Length Confounding in LLM Reasoning Data Selection","display_name":"On the Step Length Confounding in LLM Reasoning Data Selection","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166846202","doi":"https://doi.org/10.18653/v1/2026.findings-acl.918"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.918","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.918","pdf_url":"https://aclanthology.org/2026.findings-acl.918.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.918.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139752252","display_name":"Bing Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bing Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036927614","display_name":"Rui Miao","orcid":"https://orcid.org/0000-0002-2917-2311"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rui Miao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139834999","display_name":"Chen Shen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen Shen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050698074","display_name":"Shaotian Yan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shaotian Yan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139743090","display_name":"Kaiyuan Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kaiyuan Liu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100751173","display_name":"Ximing Li","orcid":"https://orcid.org/0000-0001-8190-5087"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ximing Li","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101377833","display_name":"Xiaosong Yuan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiaosong Yuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030567406","display_name":"Sinan Fan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sinan Fan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139834593","display_name":"Jun Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jun Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139846511","display_name":"Jieping Ye","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jieping Ye","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.85237382,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"18443","last_page":"18457"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10538","display_name":"Data Mining Algorithms and Applications","score":0.12330000102519989,"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/T10538","display_name":"Data Mining Algorithms and Applications","score":0.12330000102519989,"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/T11010","display_name":"Logic, Reasoning, and Knowledge","score":0.09179999679327011,"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/T11063","display_name":"Rough Sets and Fuzzy Logic","score":0.08940000087022781,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/selection","display_name":"Selection (genetic algorithm)","score":0.4851999878883362},{"id":"https://openalex.org/keywords/feature-selection","display_name":"Feature selection","score":0.3547999858856201},{"id":"https://openalex.org/keywords/confounding","display_name":"Confounding","score":0.2754000127315521},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.26930001378059387}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5644000172615051},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5296000242233276},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.4851999878883362},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3628000020980835},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.3547999858856201},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3450999855995178},{"id":"https://openalex.org/C77350462","wikidata":"https://www.wikidata.org/wiki/Q1125472","display_name":"Confounding","level":2,"score":0.2754000127315521},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2721000015735626},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.26930001378059387},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.24500000476837158}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.918","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.918","pdf_url":"https://aclanthology.org/2026.findings-acl.918.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.918","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.918","pdf_url":"https://aclanthology.org/2026.findings-acl.918.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G8897825800","display_name":null,"funder_award_id":"62276113","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166846202.pdf","grobid_xml":"https://content.openalex.org/works/W7166846202.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0,36],"reasoning":[1,31,78,86,115],"models":[2],"have":[3],"recently":[4],"demonstrated":[5],"strong":[6],"performance":[7],"on":[8,19],"complex":[9],"tasks":[10],"that":[11],"require":[12],"long":[13,30],"chain-of-thought":[14],"reasoning,":[15],"through":[16],"supervised":[17],"fine-tuning":[18],"large-scale":[20],"and":[21,40,146,164],"high-quality":[22,50],"datasets.To":[23],"construct":[24],"such":[25],"datasets,":[26,79],"existing":[27],"pipelines":[28],"generate":[29],"data":[32,57,61],"from":[33],"more":[34,89],"capable":[35],"Language":[37],"Models":[38],"(LLMs)":[39],"apply":[41],"manually":[42],"heuristic":[43],"or":[44],"naturalness-based":[45,56],"selection":[46],"methods":[47],"to":[48,76,110,154],"filter":[49],"samples.Despite":[51],"the":[52,63,124,156,169,176],"proven":[53],"effectiveness":[54,170],"of":[55,171],"selection,":[58],"which":[59,137,148],"ranks":[60],"by":[62,68],"average":[64,125,143],"log":[65,126,144],"probability":[66],"assigned":[67],"LLMs,":[69],"our":[70,172],"analysis":[71],"shows":[72],"that,":[73],"when":[74,141],"applied":[75],"LLM":[77],"it":[80],"systematically":[81],"prefers":[82],"samples":[83],"with":[84],"longer":[85,117],"steps":[87,118],"(i.e.,":[88],"tokens":[90,113],"per":[91],"step)":[92],"rather":[93],"than":[94],"higher-quality":[95],"ones,":[96],"a":[97,150],"phenomenon":[98,109],"we":[99,106,131],"term":[100],"step":[101,177],"length":[102,178],"confounding.Through":[103],"quantitative":[104],"analysis,":[105],"attribute":[107],"this":[108,129],"low-probability":[111],"first":[112,157],"in":[114,174],"steps;":[116],"dilute":[119],"their":[120],"influence,":[121],"thereby":[122],"inflating":[123],"probabilities.To":[127],"address":[128],"issue,":[130],"propose":[132],"two":[133],"variant":[134],"methods:":[135],"ASLEC-DROP,":[136],"drops":[138],"first-token":[139],"probabilities":[140],"computing":[142],"probability,":[145],"ASLEC-CASL,":[147],"applies":[149],"causal":[151],"debiasing":[152],"regression":[153],"remove":[155],"tokens'":[158],"confounding":[159,179],"effect.Experiments":[160],"across":[161],"four":[162],"LLMs":[163],"five":[165],"evaluation":[166],"benchmarks":[167],"demonstrate":[168],"approach":[173],"mitigating":[175],"problem.":[180]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}