{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:QTWK7M5EILECIDJ6DG5ADQ4BWN","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"981b5145510893e2148daef694e1c6fd5367bd108580836d118654d675db943e","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T04:47:16Z","title_canon_sha256":"c114c7238e8a5777a35d40442699e25cdaed58bccc3ff77b75800c8d1e103f12"},"schema_version":"1.0","source":{"id":"2605.14365","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14365","created_at":"2026-05-17T23:39:07Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14365v1","created_at":"2026-05-17T23:39:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14365","created_at":"2026-05-17T23:39:07Z"},{"alias_kind":"pith_short_12","alias_value":"QTWK7M5EILEC","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"QTWK7M5EILECIDJ6","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"QTWK7M5E","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:68addd7cf425606b24ce30ff5b0b6612dd8f8aa62e44299cd3a34546ef4d59c9","target":"graph","created_at":"2026-05-17T23:39:07Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"We prove that for r >= 2 this generalization strictly enlarges BatchEnsemble's hypothesis class. Empirically, LoMETab sustains higher pairwise KL than an additive low-rank ablation, and (r, sigma_init) provides broad control over pairwise KL, varying by up to several orders of magnitude."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the added representational capacity and induced diversity will translate into practically useful predictive behavior rather than merely dataset-dependent variation; the abstract reports that performance over the (r, sigma_init) grid is dataset-dependent."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"LoMETab is a rank-r generalization of multiplicative implicit ensembles that strictly enlarges the hypothesis class for r >= 2 and supplies tunable control over predictive diversity via adapter rank and initialization scale."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"LoMETab generalizes rank-1 multiplicative ensembles to rank-r adapters for tabular models, strictly enlarging the hypothesis class and exposing tunable diversity controls."}],"snapshot_sha256":"5d5426ace85027eeaa4289c21f1c5e1837e669502f77a42df925d3fe7a222bb0"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ebde908e23afe8710f643e2ac4f28152414efdc0a3796f700efb496812ff706b"},"paper":{"abstract_excerpt":"Recent tabular learning benchmarks increasingly show a tight performance cluster rather than a clear hierarchy among leading methods, spanning gradient boosted decision trees, attention-based architectures, and implicit ensembles such as TabM. As benchmark gains plateau, a complementary goal is to understand and control the mechanisms that make simple neural tabular models competitive. We propose LoMETab, a rank-$r$ generalization of multiplicative implicit ensembles. LoMETab lifts the rank-1 BatchEnsemble/TabM modulation to a rank-$r$ identity-residual Hadamard family by parameterizing each m","authors_text":"Changryeol Choi, Gowun Jeong, Hyewon Park, Yujin Kwon","cross_cats":["cs.AI"],"headline":"LoMETab generalizes rank-1 multiplicative ensembles to rank-r adapters for tabular models, strictly enlarging the hypothesis class and exposing tunable diversity controls.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T04:47:16Z","title":"LoMETab: Beyond Rank-1 Ensembles for Tabular Deep Learning"},"references":{"count":23,"internal_anchors":0,"resolved_work":23,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Optuna: A next-generation hyperparameter optimization framework","work_id":"f11f4cd0-9ce5-4174-a976-f401f5e9ce28","year":2019},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"XGBoost: A scalable tree boosting system","work_id":"bea565cd-2b99-41c7-8085-02eba78c6f8f","year":2016},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Masksembles for uncertainty estimation","work_id":"c8d2a3d6-20d5-47ec-955a-24064d06b6d3","year":2021},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"On embeddings for numerical features in tabular deep learning","work_id":"aaa16ab9-7e1e-4cca-903a-566c4b099f74","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"TabR: Tabular deep learning meets nearest neighbors","work_id":"063c1227-1e8a-4626-9d08-87eb984116d0","year":2024}],"snapshot_sha256":"1cb42c6ab57edb06565febc01d19bb8e642782881483b3ce025f6df37199a497"},"source":{"id":"2605.14365","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-15T02:49:04.351626Z","id":"fb9c7d3c-d33a-4766-8274-0424ae153871","model_set":{"reader":"grok-4.3"},"one_line_summary":"LoMETab is a rank-r generalization of multiplicative implicit ensembles that strictly enlarges the hypothesis class for r >= 2 and supplies tunable control over predictive diversity via adapter rank and initialization scale.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"LoMETab generalizes rank-1 multiplicative ensembles to rank-r adapters for tabular models, strictly enlarging the hypothesis class and exposing tunable diversity controls.","strongest_claim":"We prove that for r >= 2 this generalization strictly enlarges BatchEnsemble's hypothesis class. Empirically, LoMETab sustains higher pairwise KL than an additive low-rank ablation, and (r, sigma_init) provides broad control over pairwise KL, varying by up to several orders of magnitude.","weakest_assumption":"That the added representational capacity and induced diversity will translate into practically useful predictive behavior rather than merely dataset-dependent variation; the abstract reports that performance over the (r, sigma_init) grid is dataset-dependent."}},"verdict_id":"fb9c7d3c-d33a-4766-8274-0424ae153871"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:80b81eb3e5b3a59589dd03d9227a89c41d6b8aef9bb1e70bedf494c296dcdd06","target":"record","created_at":"2026-05-17T23:39:07Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"981b5145510893e2148daef694e1c6fd5367bd108580836d118654d675db943e","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T04:47:16Z","title_canon_sha256":"c114c7238e8a5777a35d40442699e25cdaed58bccc3ff77b75800c8d1e103f12"},"schema_version":"1.0","source":{"id":"2605.14365","kind":"arxiv","version":1}},"canonical_sha256":"84ecafb3a442c8240d3e19ba01c381b3403c2a41856742679b94a189962e09d8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"84ecafb3a442c8240d3e19ba01c381b3403c2a41856742679b94a189962e09d8","first_computed_at":"2026-05-17T23:39:07.897817Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:07.897817Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"bWFq2VaQQPpeGlu/4u/XqIcBXZPRza4/ZpU7pqhLv9dHE4aS5WylAPZPlf/GigpcaFu/ANq69ji6dAtsqgywDg==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:07.898523Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.14365","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:80b81eb3e5b3a59589dd03d9227a89c41d6b8aef9bb1e70bedf494c296dcdd06","sha256:68addd7cf425606b24ce30ff5b0b6612dd8f8aa62e44299cd3a34546ef4d59c9"],"state_sha256":"83852784cb2a70c0ea54752bbaab3e2842deb32dc1e6ad74feff5751b9551566"}