{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:27QZILTGV2QNFYS7MQTLY3ADEV","short_pith_number":"pith:27QZILTG","schema_version":"1.0","canonical_sha256":"d7e1942e66aea0d2e25f6426bc6c032551947134a20cf4932f6f5ee4ea87dc7e","source":{"kind":"arxiv","id":"2508.11176","version":1},"attestation_state":"computed","paper":{"title":"Fine-Grained VLM Fine-tuning via Latent Hierarchical Adapter Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bin Luo, Bo Jiang, Jin Tang, Xiao Wang, Yuhe Ding, Yumiao Zhao","submitted_at":"2025-08-15T03:02:36Z","abstract_excerpt":"Adapter-based approaches have garnered attention for fine-tuning pre-trained Vision-Language Models (VLMs) on few-shot classification tasks. These methods strive to develop a lightweight module that better aligns visual and (category) textual representations, thereby enhancing performance on downstream few-shot learning tasks. However, existing adapters generally learn/align (category) textual-visual modalities via explicit spatial proximity in the underlying embedding space, which i) fails to capture the inherent one-to-many associations between categories and image samples and ii) struggles "},"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":"2508.11176","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-08-15T03:02:36Z","cross_cats_sorted":[],"title_canon_sha256":"eae9f78e3b81bfb0be8ee0b0c1f723d3131e98e2cad4feefc61c7d86930896ff","abstract_canon_sha256":"29cca52a9f5eee8a4c022cea80b60ccd9654dd9b2f6abfea477b586ab65e9247"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:54:19.920890Z","signature_b64":"LhZQ++dcuTuUPXcvfXPw6AN9VkIdlxVd8XW26/2T3O3kOZhYlmfLColWyWs7W0T7U39YN0mi9B+Cuhnh7x1lDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d7e1942e66aea0d2e25f6426bc6c032551947134a20cf4932f6f5ee4ea87dc7e","last_reissued_at":"2026-07-05T11:54:19.920430Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:54:19.920430Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fine-Grained VLM Fine-tuning via Latent Hierarchical Adapter Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bin Luo, Bo Jiang, Jin Tang, Xiao Wang, Yuhe Ding, Yumiao Zhao","submitted_at":"2025-08-15T03:02:36Z","abstract_excerpt":"Adapter-based approaches have garnered attention for fine-tuning pre-trained Vision-Language Models (VLMs) on few-shot classification tasks. These methods strive to develop a lightweight module that better aligns visual and (category) textual representations, thereby enhancing performance on downstream few-shot learning tasks. However, existing adapters generally learn/align (category) textual-visual modalities via explicit spatial proximity in the underlying embedding space, which i) fails to capture the inherent one-to-many associations between categories and image samples and ii) struggles "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.11176","kind":"arxiv","version":1},"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/2508.11176/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":"2508.11176","created_at":"2026-07-05T11:54:19.920483+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.11176v1","created_at":"2026-07-05T11:54:19.920483+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.11176","created_at":"2026-07-05T11:54:19.920483+00:00"},{"alias_kind":"pith_short_12","alias_value":"27QZILTGV2QN","created_at":"2026-07-05T11:54:19.920483+00:00"},{"alias_kind":"pith_short_16","alias_value":"27QZILTGV2QNFYS7","created_at":"2026-07-05T11:54:19.920483+00:00"},{"alias_kind":"pith_short_8","alias_value":"27QZILTG","created_at":"2026-07-05T11:54:19.920483+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.00428","citing_title":"HyFL-CLIP: Hyperbolic Fine-Tuning of CLIP for Robust Long-Context Understanding","ref_index":82,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28051","citing_title":"Beyond Surrogate Gradients: Fully Differentiable Token Pruning for Vision-Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09392","citing_title":"HyNeuralMap: Hyperbolic Mapping of Visual Semantics to Neural Hierarchies","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06285","citing_title":"Harnessing Hyperbolic Geometry for Harmful Prompt Detection and Sanitization","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV","json":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV.json","graph_json":"https://pith.science/api/pith-number/27QZILTGV2QNFYS7MQTLY3ADEV/graph.json","events_json":"https://pith.science/api/pith-number/27QZILTGV2QNFYS7MQTLY3ADEV/events.json","paper":"https://pith.science/paper/27QZILTG"},"agent_actions":{"view_html":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV","download_json":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV.json","view_paper":"https://pith.science/paper/27QZILTG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.11176&json=true","fetch_graph":"https://pith.science/api/pith-number/27QZILTGV2QNFYS7MQTLY3ADEV/graph.json","fetch_events":"https://pith.science/api/pith-number/27QZILTGV2QNFYS7MQTLY3ADEV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV/action/storage_attestation","attest_author":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV/action/author_attestation","sign_citation":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV/action/citation_signature","submit_replication":"https://pith.science/pith/27QZILTGV2QNFYS7MQTLY3ADEV/action/replication_record"}},"created_at":"2026-07-05T11:54:19.920483+00:00","updated_at":"2026-07-05T11:54:19.920483+00:00"}