{"id":"https://openalex.org/W7123249017","doi":"https://doi.org/10.48550/arxiv.2601.05770","title":"Weights to Code: Extracting Interpretable Algorithms from the Discrete Transformer","display_name":"Weights to Code: Extracting Interpretable Algorithms from the Discrete Transformer","publication_year":2026,"publication_date":"2026-01-09","ids":{"openalex":"https://openalex.org/W7123249017","doi":"https://doi.org/10.48550/arxiv.2601.05770"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2601.05770","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.05770","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2601.05770","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5122818881","display_name":"Yifan Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Zhang, Yifan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101671513","display_name":"Wei Bi","orcid":"https://orcid.org/0000-0002-0308-4217"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bi, Wei","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122771883","display_name":"Kechi Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Kechi","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122834039","display_name":"Dongming Jin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jin, Dongming","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122835128","display_name":"Jie Fu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fu, Jie","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5122803260","display_name":"Zhi Jin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jin, Zhi","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5122818881"],"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/T11975","display_name":"Evolutionary Algorithms and Applications","score":0.298799991607666,"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/T11975","display_name":"Evolutionary Algorithms and Applications","score":0.298799991607666,"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":0.10520000010728836,"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/T10260","display_name":"Software Engineering Research","score":0.056299999356269836,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.7103999853134155},{"id":"https://openalex.org/keywords/interpretability","display_name":"Interpretability","score":0.6495000123977661},{"id":"https://openalex.org/keywords/executable","display_name":"Executable","score":0.4772999882698059},{"id":"https://openalex.org/keywords/preprocessor","display_name":"Preprocessor","score":0.38499999046325684},{"id":"https://openalex.org/keywords/heuristics","display_name":"Heuristics","score":0.357699990272522},{"id":"https://openalex.org/keywords/symbolic-data-analysis","display_name":"Symbolic data analysis","score":0.33959999680519104}],"concepts":[{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.7103999853134155},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.6507999897003174},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.6495000123977661},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6258000135421753},{"id":"https://openalex.org/C160145156","wikidata":"https://www.wikidata.org/wiki/Q778586","display_name":"Executable","level":2,"score":0.4772999882698059},{"id":"https://openalex.org/C34736171","wikidata":"https://www.wikidata.org/wiki/Q918333","display_name":"Preprocessor","level":2,"score":0.38499999046325684},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.357699990272522},{"id":"https://openalex.org/C65620979","wikidata":"https://www.wikidata.org/wiki/Q7661176","display_name":"Symbolic data analysis","level":2,"score":0.33959999680519104},{"id":"https://openalex.org/C126980161","wikidata":"https://www.wikidata.org/wiki/Q863783","display_name":"Simulated annealing","level":2,"score":0.31929999589920044},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.28619998693466187},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.2623000144958496},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.25589999556541443},{"id":"https://openalex.org/C73000952","wikidata":"https://www.wikidata.org/wiki/Q17007827","display_name":"Discretization","level":2,"score":0.2556999921798706},{"id":"https://openalex.org/C1893757","wikidata":"https://www.wikidata.org/wiki/Q3653001","display_name":"Inversion (geology)","level":3,"score":0.25290000438690186},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25200000405311584}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2601.05770","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.05770","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2601.05770","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.05770","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Algorithm":[0],"extraction":[1],"aims":[2],"to":[3,28,59,85,96,102],"synthesize":[4],"executable":[5],"programs":[6],"directly":[7],"from":[8],"models":[9],"trained":[10],"on":[11,21],"algorithmic":[12],"tasks,":[13],"enabling":[14],"de":[15],"novo":[16],"algorithm":[17,138],"discovery":[18,139],"without":[19],"relying":[20],"human-written":[22],"code.":[23],"However,":[24],"applying":[25],"this":[26],"paradigm":[27],"Transformer":[29,92,131,141],"is":[30],"hindered":[31],"by":[32],"representation":[33],"entanglement":[34],"(e.g.,":[35],"superposition),":[36],"where":[37],"entangled":[38],"features":[39],"encoded":[40],"in":[41],"overlapping":[42],"directions":[43],"obstruct":[44],"the":[45,52,61,90,107,129],"recovery":[46],"of":[47],"symbolic":[48,68,83],"expressions.":[49],"We":[50],"propose":[51],"Discrete":[53,91,130],"Transformer,":[54],"an":[55],"architecture":[56],"explicitly":[57],"designed":[58],"bridge":[60],"gap":[62],"between":[63],"continuous":[64,103],"representations":[65],"and":[66,82,106,140],"discrete":[67],"logic.":[69],"By":[70],"injecting":[71],"discreteness":[72],"through":[73],"temperature-annealed":[74],"sampling,":[75],"our":[76],"framework":[77,135],"effectively":[78],"leverages":[79],"hypothesis":[80],"testing":[81],"regression":[84],"extract":[86],"human-readable":[87],"programs.":[88],"Empirically,":[89],"achieves":[93],"performance":[94],"comparable":[95],"RNN-based":[97],"methods":[98],"while":[99],"extending":[100],"interpretability":[101],"variable":[104],"domains,":[105],"annealing":[108],"dynamics":[109],"exhibit":[110],"a":[111,133],"clear":[112],"exploration-to-exploitation":[113],"transition.":[114],"Finally,":[115],"we":[116],"show":[117],"that":[118],"architectural":[119],"inductive":[120],"biases":[121],"provide":[122],"fine-grained":[123],"control":[124],"over":[125],"synthesized":[126],"programs,":[127],"establishing":[128],"as":[132],"robust":[134],"for":[136],"demonstration-free":[137],"interpretability.":[142]},"counts_by_year":[],"updated_date":"2026-03-21T06:30:42.041108","created_date":"2026-01-13T00:00:00"}