{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:GE2DIR2QLUPWAXS56AQ3PMYF6M","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"d56fe6ecb3a8caf23fc7cfd2ce12e7db1133bbd87c3d8c74c283c9ad2a5c5af5","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T11:27:46Z","title_canon_sha256":"3f9da7b42bbb9de9deeb4fe6777ad1d8822702621470cbc499065b556526684c"},"schema_version":"1.0","source":{"id":"2605.14709","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14709","created_at":"2026-05-17T23:38:59Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14709v1","created_at":"2026-05-17T23:38:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14709","created_at":"2026-05-17T23:38:59Z"},{"alias_kind":"pith_short_12","alias_value":"GE2DIR2QLUPW","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"GE2DIR2QLUPWAXS5","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"GE2DIR2Q","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:504abe4ea8896fbc3a4da9b6ff6a34e6eedd5bced2429144b39ba0f490ca7f66","target":"graph","created_at":"2026-05-17T23:38:59Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"our method outperforms existing baselines on X2I, achieving superior generation fidelity among simple-to-complex instructions."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The constructed hierarchical data pipeline and designed step-wise rewards plus complexity penalty will enable effective autonomous mode switching without creating new bottlenecks or overfitting to the new dataset."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Unified multimodal models gain self-adaptive modes (direct generation, self-reflection, multi-step planning) trained via SFT and RL with step-wise rewards to close the understanding-generation gap in anything-to-image tasks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Unified multimodal models learn to switch autonomously between direct generation, reflection, and planning to close the understanding-generation gap in image tasks."}],"snapshot_sha256":"19f2cdc0898cd1d8245cd4eeec758901dc0d49d86dfd748b55c96957796192d0"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Recent unified models integrate multimodal understanding and generation within a single framework. However, an \"understanding-generation gap\" persists, where models can capture user intent but often fail to translate this semantic knowledge into precise pixel-level manipulation. This gap results in two bottlenecks in anything-to-image task (X2I): the attention entanglement bottleneck, where blind planning struggles with complex prompts, and the visual refinement bottleneck, where unstructured feedback fails to correct imperfections efficiently. In this paper, we propose a novel framework that ","authors_text":"Bingjie Gao, Canmiao Fu, Chen Li, Feng Wang, Jiangtong Li, Keming Ye, Li Niu, Qingyang Liu, Shaobo Wang, Shuochen Chang, Yali Wang, Zhipeng Huang","cross_cats":[],"headline":"Unified multimodal models learn to switch autonomously between direct generation, reflection, and planning to close the understanding-generation gap in image tasks.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T11:27:46Z","title":"Breaking Dual Bottlenecks: Evolving Unified Multimodal Models into Self-Adaptive Interleaved Visual Reasoners"},"references":{"count":36,"internal_anchors":0,"resolved_work":36,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Thinking-while-generating: Interleaving tex- tual reasoning throughout visual generation.arXiv preprint arXiv:2511.16671, 2025","work_id":"4216b182-04a6-4cc1-a898-5aaf370461c3","year":2026},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Your Objective: Evaluate how faithfully the Generated Image (Y) fulfills the **Instruction**, focusing on whether the requested changes or additions were executed correctly","work_id":"75a3f053-0bdd-4da5-8561-5cc7940f7220","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"**Detect Change**: What has been added, modified, or created in Y compared to X? (If X is Text-only, evaluate Y directly against the text)","work_id":"aaacd748-0e84-4cff-92a3-310083988a91","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":6,"title":"**Expected Visual Caption**: Describe the ideal result if the instruction were perfectly followed","work_id":"bff7bb31-bdbd-461e-9c21-4e0f996534bc","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":7,"title":"**Instruction Match**: - Was the correct subject/attribute modified or created? - For **Spatial/Size** changes: Is the placement or scale correct relative to the instruction? - For **Subject-driven** ","work_id":"93309ce4-e4b7-410c-8853-3d326e277ecb","year":null}],"snapshot_sha256":"dee9141e29d50e0c38e6a73ffaf497d33e34a1180894ed3d1411082df59ead52"},"source":{"id":"2605.14709","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-15T04:47:27.286732Z","id":"ceb3a763-2059-41cb-9dd6-1a80e63d63eb","model_set":{"reader":"grok-4.3"},"one_line_summary":"Unified multimodal models gain self-adaptive modes (direct generation, self-reflection, multi-step planning) trained via SFT and RL with step-wise rewards to close the understanding-generation gap in anything-to-image tasks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Unified multimodal models learn to switch autonomously between direct generation, reflection, and planning to close the understanding-generation gap in image tasks.","strongest_claim":"our method outperforms existing baselines on X2I, achieving superior generation fidelity among simple-to-complex instructions.","weakest_assumption":"The constructed hierarchical data pipeline and designed step-wise rewards plus complexity penalty will enable effective autonomous mode switching without creating new bottlenecks or overfitting to the new dataset."}},"verdict_id":"ceb3a763-2059-41cb-9dd6-1a80e63d63eb"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:4bfb413f548d387ec3bd8ad47e03bda5b50f55125bc124ddac18a383ca7562fe","target":"record","created_at":"2026-05-17T23:38:59Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"d56fe6ecb3a8caf23fc7cfd2ce12e7db1133bbd87c3d8c74c283c9ad2a5c5af5","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T11:27:46Z","title_canon_sha256":"3f9da7b42bbb9de9deeb4fe6777ad1d8822702621470cbc499065b556526684c"},"schema_version":"1.0","source":{"id":"2605.14709","kind":"arxiv","version":1}},"canonical_sha256":"31343447505d1f605e5df021b7b305f3025e339a2b83095c51f7828270feb58b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"31343447505d1f605e5df021b7b305f3025e339a2b83095c51f7828270feb58b","first_computed_at":"2026-05-17T23:38:59.234134Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:59.234134Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"wBAKn/sYrbfR/qz38HUkvAPN6gm6stZEKQGKn1db6B3dEjvaecz5neGMnQ+JdlBpD5AT5OGDe7mIoCNFMp3dBw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:59.234645Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.14709","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4bfb413f548d387ec3bd8ad47e03bda5b50f55125bc124ddac18a383ca7562fe","sha256:504abe4ea8896fbc3a4da9b6ff6a34e6eedd5bced2429144b39ba0f490ca7f66"],"state_sha256":"e2137c4ea73ca24d8ee1abfbaa1b1499a926a2aa545082eb16d7d156ad7763a2"}