{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5IC2HMSOXRXMBO66AVVN2PX22G","short_pith_number":"pith:5IC2HMSO","schema_version":"1.0","canonical_sha256":"ea05a3b24ebc6ec0bbde056add3efad1a77207408bac2c885776722b379059ed","source":{"kind":"arxiv","id":"2606.29579","version":1},"attestation_state":"computed","paper":{"title":"ScAle: Attention Head Scaling as a Minimal Adapter for Spatial Reasoning in Vision Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MM"],"primary_cat":"cs.CV","authors_text":"Pu Zhao, Rahul Chowdhury, Timothy A Rupprecht, Xuan Shen, Yanzhi Wang","submitted_at":"2026-06-28T19:48:39Z","abstract_excerpt":"Spatial reasoning remains a persistent challenge for many vision language models (VLMs), and improving it typically requires fine-tuning with substantial additional parameters. Our preliminary analysis reveals that rescaling activations in selected transformer layers-without modifying pretrained weights-can significantly influence downstream performance. Motivated by this observation, we propose ScAle, an ultra-lightweight adaptation method that learns a small set of scalar coefficients to modulate last-token attention and MLP activations in a fully frozen backbone. We evaluate our method on t"},"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":"2606.29579","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2026-06-28T19:48:39Z","cross_cats_sorted":["cs.AI","cs.LG","cs.MM"],"title_canon_sha256":"61e4e88825aa7a0642ab8b30671a3c1284426d6dbed742ea3176e270dbda7866","abstract_canon_sha256":"c8490cc9193f8a92e510ff825695683ef218b547d50276316bb8029f623ec067"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T01:18:12.923189Z","signature_b64":"yg/kyCOnxaVOnlCwX4s1NxS59CjzQPZA5JVhLgbnKOlxZoMLEJ3LpR688YKYku+3hHSdl9lUb7HCsFfoYF0KDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ea05a3b24ebc6ec0bbde056add3efad1a77207408bac2c885776722b379059ed","last_reissued_at":"2026-06-30T01:18:12.922726Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T01:18:12.922726Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ScAle: Attention Head Scaling as a Minimal Adapter for Spatial Reasoning in Vision Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MM"],"primary_cat":"cs.CV","authors_text":"Pu Zhao, Rahul Chowdhury, Timothy A Rupprecht, Xuan Shen, Yanzhi Wang","submitted_at":"2026-06-28T19:48:39Z","abstract_excerpt":"Spatial reasoning remains a persistent challenge for many vision language models (VLMs), and improving it typically requires fine-tuning with substantial additional parameters. Our preliminary analysis reveals that rescaling activations in selected transformer layers-without modifying pretrained weights-can significantly influence downstream performance. Motivated by this observation, we propose ScAle, an ultra-lightweight adaptation method that learns a small set of scalar coefficients to modulate last-token attention and MLP activations in a fully frozen backbone. We evaluate our method on t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.29579","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/2606.29579/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":"2606.29579","created_at":"2026-06-30T01:18:12.922793+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.29579v1","created_at":"2026-06-30T01:18:12.922793+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.29579","created_at":"2026-06-30T01:18:12.922793+00:00"},{"alias_kind":"pith_short_12","alias_value":"5IC2HMSOXRXM","created_at":"2026-06-30T01:18:12.922793+00:00"},{"alias_kind":"pith_short_16","alias_value":"5IC2HMSOXRXMBO66","created_at":"2026-06-30T01:18:12.922793+00:00"},{"alias_kind":"pith_short_8","alias_value":"5IC2HMSO","created_at":"2026-06-30T01:18:12.922793+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G","json":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G.json","graph_json":"https://pith.science/api/pith-number/5IC2HMSOXRXMBO66AVVN2PX22G/graph.json","events_json":"https://pith.science/api/pith-number/5IC2HMSOXRXMBO66AVVN2PX22G/events.json","paper":"https://pith.science/paper/5IC2HMSO"},"agent_actions":{"view_html":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G","download_json":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G.json","view_paper":"https://pith.science/paper/5IC2HMSO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.29579&json=true","fetch_graph":"https://pith.science/api/pith-number/5IC2HMSOXRXMBO66AVVN2PX22G/graph.json","fetch_events":"https://pith.science/api/pith-number/5IC2HMSOXRXMBO66AVVN2PX22G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G/action/storage_attestation","attest_author":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G/action/author_attestation","sign_citation":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G/action/citation_signature","submit_replication":"https://pith.science/pith/5IC2HMSOXRXMBO66AVVN2PX22G/action/replication_record"}},"created_at":"2026-06-30T01:18:12.922793+00:00","updated_at":"2026-06-30T01:18:12.922793+00:00"}