{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UOSPVVQFIRBAPFANRR3OX2SYW6","short_pith_number":"pith:UOSPVVQF","schema_version":"1.0","canonical_sha256":"a3a4fad605444207940d8c76ebea58b7bd037b4e4892f03d82ea5dc91e10b65c","source":{"kind":"arxiv","id":"2503.08497","version":2},"attestation_state":"computed","paper":{"title":"MMRL: Multi-Modal Representation Learning for Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Xiaodong Gu, Yuncheng Guo","submitted_at":"2025-03-11T14:48:01Z","abstract_excerpt":"Large-scale pre-trained Vision-Language Models (VLMs) have become essential for transfer learning across diverse tasks. However, adapting these models with limited few-shot data often leads to overfitting, diminishing their performance on new tasks. To tackle this issue, we propose a novel Multi-Modal Representation Learning (MMRL) framework that introduces a shared, learnable, and modality-agnostic representation space. MMRL projects the space tokens to text and image representation tokens, facilitating more effective multi-modal interactions. Unlike previous approaches that solely optimize c"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2503.08497","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-03-11T14:48:01Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"1e6a1ba9ff084465724bfedcccd5167635a26d3d3e9862d2013b046ecb6b73f3","abstract_canon_sha256":"508e4b69d3f34184b4338efb711e47d20ae881fa47987254c9a2f54f2ae94614"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:39:26.888470Z","signature_b64":"LL4astGkBVw9TpeHJQuWdWv0Bzt8g5coNx1EaUvoe7brBPUm4FQfFqdiEoepyat7QAqgH2QAs6J5zq2b9zSwCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a3a4fad605444207940d8c76ebea58b7bd037b4e4892f03d82ea5dc91e10b65c","last_reissued_at":"2026-07-05T10:39:26.887929Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:39:26.887929Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MMRL: Multi-Modal Representation Learning for Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Xiaodong Gu, Yuncheng Guo","submitted_at":"2025-03-11T14:48:01Z","abstract_excerpt":"Large-scale pre-trained Vision-Language Models (VLMs) have become essential for transfer learning across diverse tasks. However, adapting these models with limited few-shot data often leads to overfitting, diminishing their performance on new tasks. To tackle this issue, we propose a novel Multi-Modal Representation Learning (MMRL) framework that introduces a shared, learnable, and modality-agnostic representation space. MMRL projects the space tokens to text and image representation tokens, facilitating more effective multi-modal interactions. Unlike previous approaches that solely optimize c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.08497","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2503.08497/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2503.08497","created_at":"2026-07-05T10:39:26.887987+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.08497v2","created_at":"2026-07-05T10:39:26.887987+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.08497","created_at":"2026-07-05T10:39:26.887987+00:00"},{"alias_kind":"pith_short_12","alias_value":"UOSPVVQFIRBA","created_at":"2026-07-05T10:39:26.887987+00:00"},{"alias_kind":"pith_short_16","alias_value":"UOSPVVQFIRBAPFAN","created_at":"2026-07-05T10:39:26.887987+00:00"},{"alias_kind":"pith_short_8","alias_value":"UOSPVVQF","created_at":"2026-07-05T10:39:26.887987+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.13161","citing_title":"A$_3$B$_2$: Adaptive Asymmetric Adapter for Alleviating Branch Bias in Vision-Language Image Classification with Few-Shot Learning","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13161","citing_title":"A$_3$B$_2$: Adaptive Asymmetric Adapter for Alleviating Branch Bias in Vision-Language Image Classification with Few-Shot Learning","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6","json":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6.json","graph_json":"https://pith.science/api/pith-number/UOSPVVQFIRBAPFANRR3OX2SYW6/graph.json","events_json":"https://pith.science/api/pith-number/UOSPVVQFIRBAPFANRR3OX2SYW6/events.json","paper":"https://pith.science/paper/UOSPVVQF"},"agent_actions":{"view_html":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6","download_json":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6.json","view_paper":"https://pith.science/paper/UOSPVVQF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.08497&json=true","fetch_graph":"https://pith.science/api/pith-number/UOSPVVQFIRBAPFANRR3OX2SYW6/graph.json","fetch_events":"https://pith.science/api/pith-number/UOSPVVQFIRBAPFANRR3OX2SYW6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6/action/storage_attestation","attest_author":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6/action/author_attestation","sign_citation":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6/action/citation_signature","submit_replication":"https://pith.science/pith/UOSPVVQFIRBAPFANRR3OX2SYW6/action/replication_record"}},"created_at":"2026-07-05T10:39:26.887987+00:00","updated_at":"2026-07-05T10:39:26.887987+00:00"}