{"id":"https://openalex.org/W7148456020","doi":"https://doi.org/10.48550/arxiv.2604.00342","title":"Is One Token All It Takes? Graph Pooling Tokens for LLM-based GraphQA","display_name":"Is One Token All It Takes? Graph Pooling Tokens for LLM-based GraphQA","publication_year":2026,"publication_date":"2026-04-01","ids":{"openalex":"https://openalex.org/W7148456020","doi":"https://doi.org/10.48550/arxiv.2604.00342"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.00342","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.00342","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.00342","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5114200072","display_name":"A. A. Grover","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Grover, Ankit","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088007626","display_name":"Lodovico Giaretta","orcid":"https://orcid.org/0000-0002-0223-8907"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Giaretta, Lodovico","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5093356418","display_name":"R\u00e9mi Bourgerie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bourgerie, R\u00e9mi","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5041411651","display_name":"\u0160ar\u016bnas Girdzijauskas","orcid":"https://orcid.org/0000-0003-4516-7317"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Girdzijauskas, Sarunas","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5114200072"],"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.9291999936103821,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.9291999936103821,"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/T10028","display_name":"Topic Modeling","score":0.041200000792741776,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.006899999920278788,"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/pooling","display_name":"Pooling","score":0.661300003528595},{"id":"https://openalex.org/keywords/graph","display_name":"Graph","score":0.44530001282691956},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.42309999465942383},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.4189999997615814},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.3695000112056732},{"id":"https://openalex.org/keywords/associative-property","display_name":"Associative property","score":0.3165000081062317}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7598000168800354},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.661300003528595},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.5508999824523926},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.44530001282691956},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.42309999465942383},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.4189999997615814},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4106000065803528},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.37770000100135803},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.3695000112056732},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.32989999651908875},{"id":"https://openalex.org/C159423971","wikidata":"https://www.wikidata.org/wiki/Q177251","display_name":"Associative property","level":2,"score":0.3165000081062317},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.310699999332428},{"id":"https://openalex.org/C180356752","wikidata":"https://www.wikidata.org/wiki/Q727035","display_name":"Adjacency matrix","level":3,"score":0.3061999976634979},{"id":"https://openalex.org/C146380142","wikidata":"https://www.wikidata.org/wiki/Q1137726","display_name":"Directed graph","level":2,"score":0.29660001397132874},{"id":"https://openalex.org/C22047676","wikidata":"https://www.wikidata.org/wiki/Q898680","display_name":"Clustering coefficient","level":3,"score":0.26429998874664307},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.2619999945163727}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.00342","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.00342","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.00342","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.00342","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":"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":{"The":[0,236],"integration":[1],"of":[2,83,96,108,148],"Graph":[3,19,187],"Neural":[4],"Networks":[5],"(GNNs)":[6],"with":[7,189,232],"Large":[8],"Language":[9],"Models":[10],"(LLMs)":[11],"has":[12],"emerged":[13],"as":[14,44,194],"a":[15,61,65,106,186,195],"promising":[16],"paradigm":[17],"for":[18,26],"Question":[20],"Answering":[21],"(GraphQA).":[22],"However,":[23],"effective":[24],"methods":[25],"encoding":[27],"complex":[28],"structural":[29],"information":[30,67],"into":[31,60,129],"the":[32,81,84,93,97,146,203,210,218],"LLM's":[33],"latent":[34],"space":[35],"remain":[36,166],"an":[37],"open":[38],"challenge.":[39],"Current":[40],"state-of-the-art":[41],"architectures,":[42],"such":[43],"G-Retriever,":[45],"typically":[46],"rely":[47],"on":[48,180],"standard":[49],"GNNs":[50],"and":[51,90,111,120,159,206],"aggressive":[52],"mean":[53],"pooling":[54,113,138],"to":[55,125,173,209],"compress":[56],"entire":[57],"graph":[58,98,127],"substructures":[59],"single":[62],"token,":[63],"creating":[64],"severe":[66],"bottleneck.":[68],"This":[69,168],"work":[70],"mitigates":[71],"this":[72],"bottleneck":[73],"by":[74],"investigating":[75],"two":[76],"orthogonal":[77],"strategies:":[78],"(1)":[79],"increasing":[80],"bandwidth":[82],"graph-to-LLM":[85],"interface":[86],"via":[87,100],"multi-token":[88],"pooling,":[89],"(2)":[91],"enhancing":[92],"semantic":[94],"quality":[95],"encoder":[99],"global":[101],"attention":[102],"mechanisms.":[103],"We":[104],"evaluate":[105],"suite":[107],"hierarchical":[109,155],"pruning":[110,160],"clustering-based":[112],"operators":[114,165],"including":[115],"Top-k,":[116],"SAGPool,":[117],"DiffPool,":[118],"MinCutPool,":[119],"Virtual":[121],"Node":[122],"Pooling":[123],"(VNPool)":[124],"project":[126],"data":[128],"multiple":[130],"learnable":[131],"tokens.":[132],"Empirically,":[133],"we":[134,183,201],"demonstrate":[135,184],"that":[136,185,217],"while":[137],"introduces":[139],"significant":[140],"instability":[141],"during":[142],"soft":[143],"prompt":[144],"tuning,":[145],"application":[147],"Low-Rank":[149],"Adaptation":[150],"(LoRA)":[151],"effectively":[152],"stabilizes":[153],"specific":[154],"projections":[156],"(notably":[157],"VNPool":[158,190],"methods),":[161],"though":[162],"dense":[163],"clustering":[164],"challenging.":[167],"stabilization":[169],"allows":[170],"compressed":[171],"representations":[172],"rival":[174],"full-graph":[175],"baselines":[176],"(achieving":[177],"~73%":[178],"Hit@1":[179],"WebQSP).":[181],"Conceptually,":[182],"Transformer":[188],"implementation":[191,237],"functions":[192],"structurally":[193],"single-layer":[196],"Perceiver":[197],"IO":[198],"encoder.":[199],"Finally,":[200],"adapt":[202],"FandE":[204],"(Features":[205],"Edges)":[207],"Score":[208],"generative":[211],"GraphQA":[212,219],"domain.":[213],"Our":[214],"analysis":[215],"reveals":[216],"benchmark":[220],"suffers":[221],"from":[222],"representational":[223],"saturation,":[224],"where":[225],"target":[226],"answers":[227],"are":[228],"often":[229],"highly":[230],"correlated":[231],"isolated":[233],"node":[234],"features.":[235],"is":[238],"available":[239],"at":[240],"https://github.com/Agrover112/G-Retriever/tree/all_good/":[241]},"counts_by_year":[],"updated_date":"2026-04-03T16:44:17.987007","created_date":"2026-04-03T00:00:00"}