{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:VTKHOGO3V4IYFX3YTK66JJ3HSR","short_pith_number":"pith:VTKHOGO3","schema_version":"1.0","canonical_sha256":"acd47719dbaf1182df789abde4a767946e845ef662d272806da732ec3fa6f84a","source":{"kind":"arxiv","id":"2410.19732","version":2},"attestation_state":"computed","paper":{"title":"Rethinking Visual Dependency in Long-Context Reasoning for Large Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Jianbing Shen, Jun Wan, Yucheng Zhou, Zhi Rao","submitted_at":"2024-10-25T17:59:09Z","abstract_excerpt":"Large Vision-Language Models (LVLMs) excel in cross-model tasks but experience performance declines in long-context reasoning due to overreliance on textual information and reduced visual dependency. In this study, we empirically analyze LVLMs in long-context reasoning, revealing that increased context length leads to a higher dependence on language at the expense of visual dependency. To address this issue, we propose a novel training-free context pruning method that selectively removes less critical textual information. Our approach enhances visual dependency and reduces textual noise, there"},"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":"2410.19732","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-25T17:59:09Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"f62ba6488863f3127af208c16f6df0d22029fddf6c8d96948455864dfb75732b","abstract_canon_sha256":"5ea76ddd5840cb5fc6a41a020213f75f0945971c0938aea29e90495419e8af28"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:52:17.519974Z","signature_b64":"6gdKVOUPr+MQkWHk+mahsHBV975YR/4fu+QfAZkIfIKtv/zrLCAs8/3UDkn0sN6e75wVtErG7yg3bVo0w4yzAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"acd47719dbaf1182df789abde4a767946e845ef662d272806da732ec3fa6f84a","last_reissued_at":"2026-07-05T09:52:17.519457Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:52:17.519457Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethinking Visual Dependency in Long-Context Reasoning for Large Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Jianbing Shen, Jun Wan, Yucheng Zhou, Zhi Rao","submitted_at":"2024-10-25T17:59:09Z","abstract_excerpt":"Large Vision-Language Models (LVLMs) excel in cross-model tasks but experience performance declines in long-context reasoning due to overreliance on textual information and reduced visual dependency. In this study, we empirically analyze LVLMs in long-context reasoning, revealing that increased context length leads to a higher dependence on language at the expense of visual dependency. To address this issue, we propose a novel training-free context pruning method that selectively removes less critical textual information. Our approach enhances visual dependency and reduces textual noise, there"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.19732","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/2410.19732/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":"2410.19732","created_at":"2026-07-05T09:52:17.519522+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.19732v2","created_at":"2026-07-05T09:52:17.519522+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.19732","created_at":"2026-07-05T09:52:17.519522+00:00"},{"alias_kind":"pith_short_12","alias_value":"VTKHOGO3V4IY","created_at":"2026-07-05T09:52:17.519522+00:00"},{"alias_kind":"pith_short_16","alias_value":"VTKHOGO3V4IYFX3Y","created_at":"2026-07-05T09:52:17.519522+00:00"},{"alias_kind":"pith_short_8","alias_value":"VTKHOGO3","created_at":"2026-07-05T09:52:17.519522+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26100","citing_title":"HierBias: Context-Conditioned Hierarchical Media Bias Detection with Multi-Task Type Classification","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19857","citing_title":"Rethinking Reinforcement Fine-Tuning in LVLM: Convergence, Reward Decomposition, and Generalization","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR","json":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR.json","graph_json":"https://pith.science/api/pith-number/VTKHOGO3V4IYFX3YTK66JJ3HSR/graph.json","events_json":"https://pith.science/api/pith-number/VTKHOGO3V4IYFX3YTK66JJ3HSR/events.json","paper":"https://pith.science/paper/VTKHOGO3"},"agent_actions":{"view_html":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR","download_json":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR.json","view_paper":"https://pith.science/paper/VTKHOGO3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.19732&json=true","fetch_graph":"https://pith.science/api/pith-number/VTKHOGO3V4IYFX3YTK66JJ3HSR/graph.json","fetch_events":"https://pith.science/api/pith-number/VTKHOGO3V4IYFX3YTK66JJ3HSR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR/action/storage_attestation","attest_author":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR/action/author_attestation","sign_citation":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR/action/citation_signature","submit_replication":"https://pith.science/pith/VTKHOGO3V4IYFX3YTK66JJ3HSR/action/replication_record"}},"created_at":"2026-07-05T09:52:17.519522+00:00","updated_at":"2026-07-05T09:52:17.519522+00:00"}