{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BSOFX3A6X5KKL4YFSE4Q6AR46H","short_pith_number":"pith:BSOFX3A6","schema_version":"1.0","canonical_sha256":"0c9c5bec1ebf54a5f30591390f023cf1f53cf4636016ae824842e0b2d5b77a97","source":{"kind":"arxiv","id":"2605.26460","version":1},"attestation_state":"computed","paper":{"title":"AnchorDiff: Training-Free Concept Grounding for MM-DiTs via Anchor-Based Graph Propagation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Jian Zhang, Zhijun Zhang","submitted_at":"2026-05-26T02:17:29Z","abstract_excerpt":"Multi-Modal Diffusion Transformers (MM-DiTs) encode rich representations for training-free concept grounding, but existing attention-based methods often produce overlapping activations on visually confusable concepts, a failure mode we call concept leakage, where target responses spill over to non-target objects. To address this issue, we propose AnchorDiff, a training-free grounding method that decouples semantic localization from structural refinement. AnchorDiff selects a high-confidence anchor from concept-to-image attention map and propagates it as a one-hot seed over a hybrid graph deriv"},"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":"2605.26460","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-26T02:17:29Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"990608f7036503d1cf03ceffbf790ae82e6c880a9620041f20001cc74738822d","abstract_canon_sha256":"01dedf6de8fdae37c21809f3dec77a329a29841cc5b0ce2e13c5e899c12b7f2b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:05:19.147230Z","signature_b64":"H8bERoO2qP9RUlfmxangMftG1urUjIZ3E4ol7i2Kq8XOg1PVrxWHk+3cpghnAYGcDiQAAXGGBWwHr9phw+NmCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0c9c5bec1ebf54a5f30591390f023cf1f53cf4636016ae824842e0b2d5b77a97","last_reissued_at":"2026-05-27T01:05:19.146466Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:05:19.146466Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AnchorDiff: Training-Free Concept Grounding for MM-DiTs via Anchor-Based Graph Propagation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Jian Zhang, Zhijun Zhang","submitted_at":"2026-05-26T02:17:29Z","abstract_excerpt":"Multi-Modal Diffusion Transformers (MM-DiTs) encode rich representations for training-free concept grounding, but existing attention-based methods often produce overlapping activations on visually confusable concepts, a failure mode we call concept leakage, where target responses spill over to non-target objects. To address this issue, we propose AnchorDiff, a training-free grounding method that decouples semantic localization from structural refinement. AnchorDiff selects a high-confidence anchor from concept-to-image attention map and propagates it as a one-hot seed over a hybrid graph deriv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.26460","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/2605.26460/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":"2605.26460","created_at":"2026-05-27T01:05:19.146596+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.26460v1","created_at":"2026-05-27T01:05:19.146596+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.26460","created_at":"2026-05-27T01:05:19.146596+00:00"},{"alias_kind":"pith_short_12","alias_value":"BSOFX3A6X5KK","created_at":"2026-05-27T01:05:19.146596+00:00"},{"alias_kind":"pith_short_16","alias_value":"BSOFX3A6X5KKL4YF","created_at":"2026-05-27T01:05:19.146596+00:00"},{"alias_kind":"pith_short_8","alias_value":"BSOFX3A6","created_at":"2026-05-27T01:05:19.146596+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/BSOFX3A6X5KKL4YFSE4Q6AR46H","json":"https://pith.science/pith/BSOFX3A6X5KKL4YFSE4Q6AR46H.json","graph_json":"https://pith.science/api/pith-number/BSOFX3A6X5KKL4YFSE4Q6AR46H/graph.json","events_json":"https://pith.science/api/pith-number/BSOFX3A6X5KKL4YFSE4Q6AR46H/events.json","paper":"https://pith.science/paper/BSOFX3A6"},"agent_actions":{"view_html":"https://pith.science/pith/BSOFX3A6X5KKL4YFSE4Q6AR46H","download_json":"https://pith.science/pith/BSOFX3A6X5KKL4YFSE4Q6AR46H.json","view_paper":"https://pith.science/paper/BSOFX3A6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.26460&json=true","fetch_graph":"https://pith.science/api/pith-number/BSOFX3A6X5KKL4YFSE4Q6AR46H/graph.json","fetch_events":"https://pith.science/api/pith-number/BSOFX3A6X5KKL4YFSE4Q6AR46H/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BSOFX3A6X5KKL4YFSE4Q6AR46H/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BSOFX3A6X5KKL4YFSE4Q6AR46H/action/storage_attestation","attest_author":"https://pith.science/pith/BSOFX3A6X5KKL4YFSE4Q6AR46H/action/author_attestation","sign_citation":"https://pith.science/pith/BSOFX3A6X5KKL4YFSE4Q6AR46H/action/citation_signature","submit_replication":"https://pith.science/pith/BSOFX3A6X5KKL4YFSE4Q6AR46H/action/replication_record"}},"created_at":"2026-05-27T01:05:19.146596+00:00","updated_at":"2026-05-27T01:05:19.146596+00:00"}