{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/re2g-retrieve-rerank-generate-2","title":"Re2G: Retrieve, Rerank, Generate","arxiv_id":"2207.06300","date":"2022-07-13","proceeding":"NAACL 2022 7","authors":["Michael Glass","Gaetano Rossiello","Md Faisal Mahbub Chowdhury","Ankita Rajaram Naik","Pengshan Cai","Alfio Gliozzo"],"abstract":"As demonstrated by GPT-3 and T5, transformers grow in capability as parameter spaces become larger and larger. However, for tasks that require a large amount of knowledge, non-parametric memory allows models to grow dramatically with a sub-linear increase in computational cost and GPU memory requirements. Recent models such as RAG and REALM have introduced retrieval into conditional generation. These models incorporate neural initial retrieval from a corpus of passages. We build on this line of research, proposing Re2G, which combines both neural initial retrieval and reranking into a BART-based sequence-to-sequence generation. Our reranking approach also permits merging retrieval results from sources with incomparable scores, enabling an ensemble of BM25 and neural initial retrieval. To train our system end-to-end, we introduce a novel variation of knowledge distillation to train the initial retrieval, reranker, and generation using only ground truth on the target sequence output. We find large gains in four diverse tasks: zero-shot slot filling, question answering, fact-checking, and dialog, with relative gains of 9% to 34% over the previous state-of-the-art on the KILT leaderboard. We make our code available as open source at https://github.com/IBM/kgi-slot-filling/tree/re2g.","url_abs":"https://arxiv.org/abs/2207.06300v1","url_pdf":"https://arxiv.org/pdf/2207.06300v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"re2g-retrieve-rerank-generate-2","repo_url":"https://github.com/ibm/kgi-slot-filling","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"fact-checking","task_name":"Fact Checking"},{"task_slug":"fact-verification","task_name":"Fact Verification"},{"task_slug":null,"task_name":"GPU"},{"task_slug":"knowledge-distillation","task_name":"Knowledge Distillation"},{"task_slug":"open-domain-dialog","task_name":"Open-Domain Dialog"},{"task_slug":"open-domain-question-answering","task_name":"Open-Domain Question Answering"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"rag","task_name":"RAG"},{"task_slug":"reranking","task_name":"Reranking"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"slot-filling","task_name":"Slot Filling"},{"task_slug":"zero-shot-slot-filling","task_name":"Zero-shot Slot Filling"}],"methods":[{"method_slug":"adafactor","method_name":"Adafactor"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"attention-dropout","method_name":"Attention Dropout"},{"method_slug":"bart","method_name":"BART"},{"method_slug":"bert","method_name":"BERT"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"cosine-annealing","method_name":"Cosine Annealing"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"gpt-3","method_name":"GPT-3"},{"method_slug":"glu","method_name":"Gated Linear Unit"},{"method_slug":"inverse-square-root-schedule","method_name":"Inverse Square Root Schedule"},{"method_slug":"knowledge-distillation","method_name":"Knowledge Distillation"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"linear-warmup-with-cosine-annealing","method_name":"Linear Warmup With Cosine Annealing"},{"method_slug":"linear-warmup-with-linear-decay","method_name":"Linear Warmup With Linear Decay"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"rag","method_name":"RAG"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"sentencepiece","method_name":"SentencePiece"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"t5","method_name":"T5"},{"method_slug":"weight-decay","method_name":"Weight Decay"},{"method_slug":"wordpiece","method_name":"WordPiece"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/fact-verification-on-kilt-fever","task":"Fact Verification","dataset":"KILT: FEVER","model":"Re2G","rank_in_archive_order":1,"of":33,"metrics":{"Accuracy":"89.55","KILT-AC":"78.53","R-Prec":"88.92","Recall@5":"92.52"},"uses_additional_data":false},{"leaderboard":"/sota/open-domain-dialog-on-kilt-wizard-of","task":"Open-Domain Dialog","dataset":"KILT: Wizard of Wikipedia","model":"Re2G","rank_in_archive_order":2,"of":21,"metrics":{"F1":"18.9","KILT-F1":"12.98","KILT-RL":"11.39","R-Prec":"60.1","ROUGE-L":"16.76","Recall@5":"79.98"},"uses_additional_data":false},{"leaderboard":"/sota/open-domain-question-answering-on-kilt","task":"Open-Domain Question Answering","dataset":"KILT: Natural Questions","model":"Re2G","rank_in_archive_order":1,"of":16,"metrics":{"EM":"51.73","F1":"60.97","KILT-EM":"43.56","KILT-F1":"49.8","R-Prec":"70.78","Recall@5":"76.63"},"uses_additional_data":false},{"leaderboard":"/sota/open-domain-question-answering-on-kilt-2","task":"Open-Domain Question Answering","dataset":"KILT: TriviaQA","model":"Re2G","rank_in_archive_order":1,"of":15,"metrics":{"EM":"76.27","F1":"81.4","KILT-EM":"57.91","KILT-F1":"61.78","R-Prec":"72.68","Recall@5":"74.23"},"uses_additional_data":false},{"leaderboard":"/sota/slot-filling-on-kilt-t-rex","task":"Slot Filling","dataset":"KILT: T-REx","model":"Re2G","rank_in_archive_order":1,"of":20,"metrics":{"Accuracy":"87.68","F1":"89.93","KILT-AC":"75.84","KILT-F1":"77.05","R-Prec":"80.7","Recall@5":"89.0"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2207.06300","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2207.06300"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ibm/kgi-slot-filling","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran_draft_wrong":1,"unverified":7},"by_repo_kind":{"official":{"samples":8,"ran":1,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"e7e75981cb464788","entry":"normalize_answer","repo":"ibm/kgi-slot-filling","repo_kind":"official","path":"slot_filling/rag_util.py","file_url":"https://github.com/ibm/kgi-slot-filling/blob/HEAD/slot_filling/rag_util.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e7e75981cb464788"}},{"code_sha256_prefix":"d890771bd3e4b3f7","entry":"clean_text","repo":"ibm/kgi-slot-filling","repo_kind":"official","path":"slot_filling/kilt_passage_corpus.py","file_url":"https://github.com/ibm/kgi-slot-filling/blob/HEAD/slot_filling/kilt_passage_corpus.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d890771bd3e4b3f7"}},{"code_sha256_prefix":"50ec835b3753c6ec","entry":"find_para_range","repo":"ibm/kgi-slot-filling","repo_kind":"official","path":"slot_filling/kilt_passage_corpus.py","file_url":"https://github.com/ibm/kgi-slot-filling/blob/HEAD/slot_filling/kilt_passage_corpus.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"50ec835b3753c6ec"}},{"code_sha256_prefix":"e57ce1467b8c1988","entry":"get_relation_from_inst","repo":"ibm/kgi-slot-filling","repo_kind":"official","path":"slot_filling/dataset_stats.py","file_url":"https://github.com/ibm/kgi-slot-filling/blob/HEAD/slot_filling/dataset_stats.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e57ce1467b8c1988"}},{"code_sha256_prefix":"0fa853d294de8b23","entry":"matching_passage_ids","repo":"ibm/kgi-slot-filling","repo_kind":"official","path":"dpr/dpr_kilt_slot_filling_dataset.py","file_url":"https://github.com/ibm/kgi-slot-filling/blob/HEAD/dpr/dpr_kilt_slot_filling_dataset.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"0fa853d294de8b23"}},{"code_sha256_prefix":"89afe8d9e930bb10","entry":"postprocess_docs","repo":"ibm/kgi-slot-filling","repo_kind":"official","path":"slot_filling/rag_util.py","file_url":"https://github.com/ibm/kgi-slot-filling/blob/HEAD/slot_filling/rag_util.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"89afe8d9e930bb10"}},{"code_sha256_prefix":"1d4a7d4d6e3eb7e0","entry":"prefix_allowed_tokens_fn","repo":"ibm/kgi-slot-filling","repo_kind":"official","path":"slot_filling/rag_client_server_apply.py","file_url":"https://github.com/ibm/kgi-slot-filling/blob/HEAD/slot_filling/rag_client_server_apply.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"1d4a7d4d6e3eb7e0"}},{"code_sha256_prefix":"f92394bf8b4685cc","entry":"prepare_seq2seq_batch","repo":"ibm/kgi-slot-filling","repo_kind":"official","path":"slot_filling/rag_util.py","file_url":"https://github.com/ibm/kgi-slot-filling/blob/HEAD/slot_filling/rag_util.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"f92394bf8b4685cc"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}