{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:DE5HB5NEEANUS4WY2X3QLHYLFJ","short_pith_number":"pith:DE5HB5NE","canonical_record":{"source":{"id":"2503.06520","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T08:48:51Z","cross_cats_sorted":["cs.MM"],"title_canon_sha256":"59dc4ae3867082357577c8d924ebc5cfbdaa49794b05dbc0ca9b5cb5270779ec","abstract_canon_sha256":"75113e3d41151dbc49215b079d83c571ce82e218aa1668b01c8d75b710268a29"},"schema_version":"1.0"},"canonical_sha256":"193a70f5a4201b4972d8d5f7059f0b2a766b9922fa756479a597657115b20c1b","source":{"kind":"arxiv","id":"2503.06520","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.06520","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"arxiv_version","alias_value":"2503.06520v2","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.06520","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"pith_short_12","alias_value":"DE5HB5NEEANU","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"DE5HB5NEEANUS4WY","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"DE5HB5NE","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:DE5HB5NEEANUS4WY2X3QLHYLFJ","target":"record","payload":{"canonical_record":{"source":{"id":"2503.06520","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T08:48:51Z","cross_cats_sorted":["cs.MM"],"title_canon_sha256":"59dc4ae3867082357577c8d924ebc5cfbdaa49794b05dbc0ca9b5cb5270779ec","abstract_canon_sha256":"75113e3d41151dbc49215b079d83c571ce82e218aa1668b01c8d75b710268a29"},"schema_version":"1.0"},"canonical_sha256":"193a70f5a4201b4972d8d5f7059f0b2a766b9922fa756479a597657115b20c1b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:47.875410Z","signature_b64":"/XKHaFLAv6/z6PhcUUyQBJKLR1qw42GKvejfXVvKHjaQ4ySnAmHeQjZ+wPhe8mkNxU9hOEuGScElreXYQ+LbCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"193a70f5a4201b4972d8d5f7059f0b2a766b9922fa756479a597657115b20c1b","last_reissued_at":"2026-05-17T23:38:47.874901Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:47.874901Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.06520","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:38:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZmR9N5+DH9GH0EdwWRtx8YHkiD2Qh+Ot/6L/V7ilrDrnrY0igbSJnNb/GGqeSFt0OGiKzf88PbI0cLDQqVd1Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T21:37:31.469461Z"},"content_sha256":"6b410cae89fb67776eb6614398d83bdc8d1cd8e70001a6b91ed2426d3b553d73","schema_version":"1.0","event_id":"sha256:6b410cae89fb67776eb6614398d83bdc8d1cd8e70001a6b91ed2426d3b553d73"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:DE5HB5NEEANUS4WY2X3QLHYLFJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Seg-Zero: Reasoning-Chain Guided Segmentation via Cognitive Reinforcement","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Reinforcement learning with format and accuracy rewards enables explicit reasoning chains to guide image segmentation.","cross_cats":["cs.MM"],"primary_cat":"cs.CV","authors_text":"Bei Yu, Bohao Peng, Fanbin Lu, Jiaya Jia, Yuqi Liu, Zhisheng Zhong, Zihao Yue","submitted_at":"2025-03-09T08:48:51Z","abstract_excerpt":"Traditional methods for reasoning segmentation rely on supervised fine-tuning with categorical labels and simple descriptions, limiting its out-of-domain generalization and lacking explicit reasoning processes. To address these limitations, we propose Seg-Zero, a novel framework that demonstrates remarkable generalizability and derives explicit chain-of-thought reasoning through cognitive reinforcement. Seg-Zero introduces a decoupled architecture consisting of a reasoning model and a segmentation model. The reasoning model interprets user intentions, generates explicit reasoning chains, and p"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Seg-Zero-7B achieves a zero-shot performance of 57.5 on the ReasonSeg benchmark, surpassing the prior LISA-7B by 18%.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the format-plus-accuracy reward mechanism, applied only through reinforcement learning without any explicit reasoning supervision, reliably produces useful and generalizable chain-of-thought reasoning rather than superficial patterns that happen to score well on the training distribution.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Seg-Zero uses cognitive reinforcement learning on a decoupled reasoning-plus-segmentation architecture to produce explicit reasoning chains and reach 57.5 zero-shot accuracy on ReasonSeg, beating prior supervised LISA-7B by 18%.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Reinforcement learning with format and accuracy rewards enables explicit reasoning chains to guide image segmentation.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"d4d9daea391d23569cc2e7031b0e2fd29746279130e22f10d41b8cc64f67839d"},"source":{"id":"2503.06520","kind":"arxiv","version":2},"verdict":{"id":"3f881baf-43c2-4cd6-b18e-dfa457c6dc42","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T12:27:04.945088Z","strongest_claim":"Seg-Zero-7B achieves a zero-shot performance of 57.5 on the ReasonSeg benchmark, surpassing the prior LISA-7B by 18%.","one_line_summary":"Seg-Zero uses cognitive reinforcement learning on a decoupled reasoning-plus-segmentation architecture to produce explicit reasoning chains and reach 57.5 zero-shot accuracy on ReasonSeg, beating prior supervised LISA-7B by 18%.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the format-plus-accuracy reward mechanism, applied only through reinforcement learning without any explicit reasoning supervision, reliably produces useful and generalizable chain-of-thought reasoning rather than superficial patterns that happen to score well on the training distribution.","pith_extraction_headline":"Reinforcement learning with format and accuracy rewards enables explicit reasoning chains to guide image segmentation."},"references":{"count":45,"sample":[{"doi":"","year":2017,"title":"Segnet: A deep convolutional encoder-decoder architecture for image segmentation","work_id":"c5380531-81cc-4c05-ab24-d89795bc0a27","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Qwen2.5-VL Technical Report","work_id":"69dffacb-bfe8-442d-be86-48624c60426f","ref_index":2,"cited_arxiv_id":"2502.13923","is_internal_anchor":true},{"doi":"","year":2025,"title":"One token to seg them all: Language instructed reasoning seg- mentation in videos","work_id":"ab23c2cb-5301-4a98-8234-5ffcf0ecf252","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2017,"title":"Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolu- tion, and fully connected crfs","work_id":"57309a7b-831c-447b-92aa-115c5bb5aefb","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Rethinking Atrous Convolution for Semantic Image Segmentation","work_id":"6f5d4c68-8df6-4794-b125-a10bfe8d5876","ref_index":5,"cited_arxiv_id":"1706.05587","is_internal_anchor":true}],"resolved_work":45,"snapshot_sha256":"63d13e76de97e481fcfc316ef5dca57081624cce43973d00957b82d449c616ae","internal_anchors":12},"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"},"verdict_id":"3f881baf-43c2-4cd6-b18e-dfa457c6dc42"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:38:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zOTOSvnDXQg27veLsEGMEwIFlhuqLXu/+7+mmpa34kRupoNWWqUmGc/yqglxD/mbW3RbUOTJo/T2qdPrcc/jAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T21:37:31.470542Z"},"content_sha256":"567018e65dc1ae80b20c10c703bef5e314179bddb8fcf9431c36f29bcb372a42","schema_version":"1.0","event_id":"sha256:567018e65dc1ae80b20c10c703bef5e314179bddb8fcf9431c36f29bcb372a42"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DE5HB5NEEANUS4WY2X3QLHYLFJ/bundle.json","state_url":"https://pith.science/pith/DE5HB5NEEANUS4WY2X3QLHYLFJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DE5HB5NEEANUS4WY2X3QLHYLFJ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-30T21:37:31Z","links":{"resolver":"https://pith.science/pith/DE5HB5NEEANUS4WY2X3QLHYLFJ","bundle":"https://pith.science/pith/DE5HB5NEEANUS4WY2X3QLHYLFJ/bundle.json","state":"https://pith.science/pith/DE5HB5NEEANUS4WY2X3QLHYLFJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DE5HB5NEEANUS4WY2X3QLHYLFJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:DE5HB5NEEANUS4WY2X3QLHYLFJ","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":"75113e3d41151dbc49215b079d83c571ce82e218aa1668b01c8d75b710268a29","cross_cats_sorted":["cs.MM"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T08:48:51Z","title_canon_sha256":"59dc4ae3867082357577c8d924ebc5cfbdaa49794b05dbc0ca9b5cb5270779ec"},"schema_version":"1.0","source":{"id":"2503.06520","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.06520","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"arxiv_version","alias_value":"2503.06520v2","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.06520","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"pith_short_12","alias_value":"DE5HB5NEEANU","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"DE5HB5NEEANUS4WY","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"DE5HB5NE","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:567018e65dc1ae80b20c10c703bef5e314179bddb8fcf9431c36f29bcb372a42","target":"graph","created_at":"2026-05-17T23:38:47Z","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":"Seg-Zero-7B achieves a zero-shot performance of 57.5 on the ReasonSeg benchmark, surpassing the prior LISA-7B by 18%."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the format-plus-accuracy reward mechanism, applied only through reinforcement learning without any explicit reasoning supervision, reliably produces useful and generalizable chain-of-thought reasoning rather than superficial patterns that happen to score well on the training distribution."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Seg-Zero uses cognitive reinforcement learning on a decoupled reasoning-plus-segmentation architecture to produce explicit reasoning chains and reach 57.5 zero-shot accuracy on ReasonSeg, beating prior supervised LISA-7B by 18%."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Reinforcement learning with format and accuracy rewards enables explicit reasoning chains to guide image segmentation."}],"snapshot_sha256":"d4d9daea391d23569cc2e7031b0e2fd29746279130e22f10d41b8cc64f67839d"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Traditional methods for reasoning segmentation rely on supervised fine-tuning with categorical labels and simple descriptions, limiting its out-of-domain generalization and lacking explicit reasoning processes. To address these limitations, we propose Seg-Zero, a novel framework that demonstrates remarkable generalizability and derives explicit chain-of-thought reasoning through cognitive reinforcement. Seg-Zero introduces a decoupled architecture consisting of a reasoning model and a segmentation model. The reasoning model interprets user intentions, generates explicit reasoning chains, and p","authors_text":"Bei Yu, Bohao Peng, Fanbin Lu, Jiaya Jia, Yuqi Liu, Zhisheng Zhong, Zihao Yue","cross_cats":["cs.MM"],"headline":"Reinforcement learning with format and accuracy rewards enables explicit reasoning chains to guide image segmentation.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T08:48:51Z","title":"Seg-Zero: Reasoning-Chain Guided Segmentation via Cognitive Reinforcement"},"references":{"count":45,"internal_anchors":12,"resolved_work":45,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Segnet: A deep convolutional encoder-decoder architecture for image segmentation","work_id":"c5380531-81cc-4c05-ab24-d89795bc0a27","year":2017},{"cited_arxiv_id":"2502.13923","doi":"","is_internal_anchor":true,"ref_index":2,"title":"Qwen2.5-VL Technical Report","work_id":"69dffacb-bfe8-442d-be86-48624c60426f","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"One token to seg them all: Language instructed reasoning seg- mentation in videos","work_id":"ab23c2cb-5301-4a98-8234-5ffcf0ecf252","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolu- tion, and fully connected crfs","work_id":"57309a7b-831c-447b-92aa-115c5bb5aefb","year":2017},{"cited_arxiv_id":"1706.05587","doi":"","is_internal_anchor":true,"ref_index":5,"title":"Rethinking Atrous Convolution for Semantic Image Segmentation","work_id":"6f5d4c68-8df6-4794-b125-a10bfe8d5876","year":null}],"snapshot_sha256":"63d13e76de97e481fcfc316ef5dca57081624cce43973d00957b82d449c616ae"},"source":{"id":"2503.06520","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-16T12:27:04.945088Z","id":"3f881baf-43c2-4cd6-b18e-dfa457c6dc42","model_set":{"reader":"grok-4.3"},"one_line_summary":"Seg-Zero uses cognitive reinforcement learning on a decoupled reasoning-plus-segmentation architecture to produce explicit reasoning chains and reach 57.5 zero-shot accuracy on ReasonSeg, beating prior supervised LISA-7B by 18%.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Reinforcement learning with format and accuracy rewards enables explicit reasoning chains to guide image segmentation.","strongest_claim":"Seg-Zero-7B achieves a zero-shot performance of 57.5 on the ReasonSeg benchmark, surpassing the prior LISA-7B by 18%.","weakest_assumption":"That the format-plus-accuracy reward mechanism, applied only through reinforcement learning without any explicit reasoning supervision, reliably produces useful and generalizable chain-of-thought reasoning rather than superficial patterns that happen to score well on the training distribution."}},"verdict_id":"3f881baf-43c2-4cd6-b18e-dfa457c6dc42"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:6b410cae89fb67776eb6614398d83bdc8d1cd8e70001a6b91ed2426d3b553d73","target":"record","created_at":"2026-05-17T23:38:47Z","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":"75113e3d41151dbc49215b079d83c571ce82e218aa1668b01c8d75b710268a29","cross_cats_sorted":["cs.MM"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T08:48:51Z","title_canon_sha256":"59dc4ae3867082357577c8d924ebc5cfbdaa49794b05dbc0ca9b5cb5270779ec"},"schema_version":"1.0","source":{"id":"2503.06520","kind":"arxiv","version":2}},"canonical_sha256":"193a70f5a4201b4972d8d5f7059f0b2a766b9922fa756479a597657115b20c1b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"193a70f5a4201b4972d8d5f7059f0b2a766b9922fa756479a597657115b20c1b","first_computed_at":"2026-05-17T23:38:47.874901Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:47.874901Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"/XKHaFLAv6/z6PhcUUyQBJKLR1qw42GKvejfXVvKHjaQ4ySnAmHeQjZ+wPhe8mkNxU9hOEuGScElreXYQ+LbCw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:47.875410Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.06520","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6b410cae89fb67776eb6614398d83bdc8d1cd8e70001a6b91ed2426d3b553d73","sha256:567018e65dc1ae80b20c10c703bef5e314179bddb8fcf9431c36f29bcb372a42"],"state_sha256":"748f863240c66303876c6959bf47da892bd1162a057470304cc845b7917f89ce"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CT0tDkPrBIgWFLJuD1pT+b+JXdmugkhqKL99G7MPNyWOIPYRccUG1rpMTIXzldc164lXRIWlJ7t/ZpILqV7IBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T21:37:31.475591Z","bundle_sha256":"0db46dad42bfb0d53714253e49aac08d496401d9f0a20cd03734909a88bd93a4"}}