{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3MUCY7KZLUC7334VVEKGKV6WE6","short_pith_number":"pith:3MUCY7KZ","schema_version":"1.0","canonical_sha256":"db282c7d595d05fdef95a9146557d627b1babe6af603ba0f4404c9acc17dd6bc","source":{"kind":"arxiv","id":"2608.01055","version":1},"attestation_state":"computed","paper":{"title":"Credit the Right Box: Marginal Contribution Assignment for Structured Visual Perception","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Feng Pan, Jianfei Wang, Shuai Li, Weixing Li, Xiang Wang, Xinheng Han, Yu Chen","submitted_at":"2026-08-02T07:43:09Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) are increasingly expected to solve structured perception tasks that require visual recognition, language-to-object binding, object cardinality preservation, and precisely localized grounding and segmentation outputs. However, existing group-relative reinforcement learning methods provide only response-level supervision, creating a granularity mismatch for structured multi-object prediction: a single advantage is broadcast to all tokens in a response, without distinguishing individual box contributions. To address this mismatch, we propose MCR-GRPO, a ma"},"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":"2608.01055","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-08-02T07:43:09Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c31033bc03c197cbedb6a9c98cd741d752d95057a013d037d97588a268d031a1","abstract_canon_sha256":"12ae92e6ba56e4bf7c2579e33d5f02a23590ca609faf1b20fd31bee83119dc8a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T01:56:26.933063Z","signature_b64":"7X1k9crPL2Pmc2v9NIRt7omjhtWnNIQpymuAKOCOq5fPkmScV/QS2/8+bhUZcngJWoAzti62WoWUXm1bdJlsCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db282c7d595d05fdef95a9146557d627b1babe6af603ba0f4404c9acc17dd6bc","last_reissued_at":"2026-08-04T01:56:26.931525Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T01:56:26.931525Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Credit the Right Box: Marginal Contribution Assignment for Structured Visual Perception","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Feng Pan, Jianfei Wang, Shuai Li, Weixing Li, Xiang Wang, Xinheng Han, Yu Chen","submitted_at":"2026-08-02T07:43:09Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) are increasingly expected to solve structured perception tasks that require visual recognition, language-to-object binding, object cardinality preservation, and precisely localized grounding and segmentation outputs. However, existing group-relative reinforcement learning methods provide only response-level supervision, creating a granularity mismatch for structured multi-object prediction: a single advantage is broadcast to all tokens in a response, without distinguishing individual box contributions. To address this mismatch, we propose MCR-GRPO, a ma"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01055","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/2608.01055/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":"2608.01055","created_at":"2026-08-04T01:56:26.932823+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.01055v1","created_at":"2026-08-04T01:56:26.932823+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01055","created_at":"2026-08-04T01:56:26.932823+00:00"},{"alias_kind":"pith_short_12","alias_value":"3MUCY7KZLUC7","created_at":"2026-08-04T01:56:26.932823+00:00"},{"alias_kind":"pith_short_16","alias_value":"3MUCY7KZLUC7334V","created_at":"2026-08-04T01:56:26.932823+00:00"},{"alias_kind":"pith_short_8","alias_value":"3MUCY7KZ","created_at":"2026-08-04T01:56:26.932823+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/3MUCY7KZLUC7334VVEKGKV6WE6","json":"https://pith.science/pith/3MUCY7KZLUC7334VVEKGKV6WE6.json","graph_json":"https://pith.science/api/pith-number/3MUCY7KZLUC7334VVEKGKV6WE6/graph.json","events_json":"https://pith.science/api/pith-number/3MUCY7KZLUC7334VVEKGKV6WE6/events.json","paper":"https://pith.science/paper/3MUCY7KZ"},"agent_actions":{"view_html":"https://pith.science/pith/3MUCY7KZLUC7334VVEKGKV6WE6","download_json":"https://pith.science/pith/3MUCY7KZLUC7334VVEKGKV6WE6.json","view_paper":"https://pith.science/paper/3MUCY7KZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.01055&json=true","fetch_graph":"https://pith.science/api/pith-number/3MUCY7KZLUC7334VVEKGKV6WE6/graph.json","fetch_events":"https://pith.science/api/pith-number/3MUCY7KZLUC7334VVEKGKV6WE6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3MUCY7KZLUC7334VVEKGKV6WE6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3MUCY7KZLUC7334VVEKGKV6WE6/action/storage_attestation","attest_author":"https://pith.science/pith/3MUCY7KZLUC7334VVEKGKV6WE6/action/author_attestation","sign_citation":"https://pith.science/pith/3MUCY7KZLUC7334VVEKGKV6WE6/action/citation_signature","submit_replication":"https://pith.science/pith/3MUCY7KZLUC7334VVEKGKV6WE6/action/replication_record"}},"created_at":"2026-08-04T01:56:26.932823+00:00","updated_at":"2026-08-04T01:56:26.932823+00:00"}