{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:JBUWOOUAVABPVIYMQYAFGN2BOA","short_pith_number":"pith:JBUWOOUA","schema_version":"1.0","canonical_sha256":"4869673a80a802faa30c860053374170261a6e4b192f59f66ac9e363304a2389","source":{"kind":"arxiv","id":"2504.11455","version":1},"attestation_state":"computed","paper":{"title":"SimpleAR: Pushing the Frontier of Autoregressive Visual Generation through Pretraining, SFT, and RL","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Junke Wang, Weilin Huang, Xinyu Zhang, Xun Wang, Yu-Gang Jiang, Zhi Tian, Zuxuan Wu","submitted_at":"2025-04-15T17:59:46Z","abstract_excerpt":"This work presents SimpleAR, a vanilla autoregressive visual generation framework without complex architecure modifications. Through careful exploration of training and inference optimization, we demonstrate that: 1) with only 0.5B parameters, our model can generate 1024x1024 resolution images with high fidelity, and achieve competitive results on challenging text-to-image benchmarks, e.g., 0.59 on GenEval and 79.66 on DPG; 2) both supervised fine-tuning (SFT) and Group Relative Policy Optimization (GRPO) training could lead to significant improvements on generation aesthectics and prompt alig"},"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":"2504.11455","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-04-15T17:59:46Z","cross_cats_sorted":[],"title_canon_sha256":"7cd75869ff0274ca10078f650e5626a2632944b32e3804cb75ce15b8870e4833","abstract_canon_sha256":"e0d0934c860b4e698e11a1d2da3b06125154b1d694e48bfbe2f26c24c9ac1495"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:49:34.179177Z","signature_b64":"zQxNstesQS98u4QYRkiDjtelbBlX1+8MOzf336qKQQm3+2XbeV7iJFAVgAmtfeF/Vi3O0J47+Fu1Xsp29PjGAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4869673a80a802faa30c860053374170261a6e4b192f59f66ac9e363304a2389","last_reissued_at":"2026-07-05T10:49:34.178708Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:49:34.178708Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SimpleAR: Pushing the Frontier of Autoregressive Visual Generation through Pretraining, SFT, and RL","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Junke Wang, Weilin Huang, Xinyu Zhang, Xun Wang, Yu-Gang Jiang, Zhi Tian, Zuxuan Wu","submitted_at":"2025-04-15T17:59:46Z","abstract_excerpt":"This work presents SimpleAR, a vanilla autoregressive visual generation framework without complex architecure modifications. Through careful exploration of training and inference optimization, we demonstrate that: 1) with only 0.5B parameters, our model can generate 1024x1024 resolution images with high fidelity, and achieve competitive results on challenging text-to-image benchmarks, e.g., 0.59 on GenEval and 79.66 on DPG; 2) both supervised fine-tuning (SFT) and Group Relative Policy Optimization (GRPO) training could lead to significant improvements on generation aesthectics and prompt alig"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.11455","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/2504.11455/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":"2504.11455","created_at":"2026-07-05T10:49:34.178768+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.11455v1","created_at":"2026-07-05T10:49:34.178768+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.11455","created_at":"2026-07-05T10:49:34.178768+00:00"},{"alias_kind":"pith_short_12","alias_value":"JBUWOOUAVABP","created_at":"2026-07-05T10:49:34.178768+00:00"},{"alias_kind":"pith_short_16","alias_value":"JBUWOOUAVABPVIYM","created_at":"2026-07-05T10:49:34.178768+00:00"},{"alias_kind":"pith_short_8","alias_value":"JBUWOOUA","created_at":"2026-07-05T10:49:34.178768+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":22,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.18518","citing_title":"UDM-GRPO: Stable and Efficient Group Relative Policy Optimization for Uniform Discrete Diffusion Models","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2606.25906","citing_title":"OracleAnalyser: Analysing Implicit Semantics of Oracle Bone Scripts through MLLMs with Post-training","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11096","citing_title":"IDEAL: In-DEpth ALignment Makes A Discrete Representation AutoEncoder","ref_index":54,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11188","citing_title":"ARM: An AutoRegressive Large Multimodal Model with Unified Discrete Representations","ref_index":81,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09156","citing_title":"OmniGen-AR: AutoRegressive Any-to-Image Generation","ref_index":82,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07967","citing_title":"DisCo: World Models with Discrete Camera Motion Control","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06078","citing_title":"Knowledge Distillation for Visual Autoregressive Models","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03216","citing_title":"Follow-Your-Preference++: Rethinking Preference Alignment for Image Inpainting","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14742","citing_title":"EARL: Towards a Unified Analysis-Guided Reinforcement Learning Framework for Egocentric Interaction Reasoning and Pixel Grounding","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00310","citing_title":"Where to Refine, When to Stop: Rethinking Redundancy via Latent Discrepancy for Efficient Visual Autoregressive Generation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2509.23352","citing_title":"Dynamic-TreeRPO: Breaking the Independent Trajectory Bottleneck with Structured Sampling","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2601.01593","citing_title":"Beyond Patches: Global-aware Autoregressive Model for Multimodal Few-Shot Font Generation","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21195","citing_title":"RankE: End-to-End Post-Training for Discrete Text-to-Image Generation with Decoder Co-Evolution","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2505.05472","citing_title":"Mogao: An Omni Foundation Model for Interleaved Multi-Modal Generation","ref_index":80,"is_internal_anchor":false},{"citing_arxiv_id":"2511.22663","citing_title":"AIA: Rethinking Architecture Decoupling Strategy In Unified Multimodal Model","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02355","citing_title":"From Broad Exploration to Stable Synthesis: Entropy-Guided Optimization for Autoregressive Image Generation","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2603.22844","citing_title":"PhySe-RPO: Physics and Semantics Guided Relative Policy Optimization for Diffusion-Based Surgical Smoke Removal","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09430","citing_title":"FlashAR: Efficient Post-Training Acceleration for Autoregressive Image Generation","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12305","citing_title":"Images in Sentences: Scaling Interleaved Instructions for Unified Visual Generation","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09430","citing_title":"FlashAR: Efficient Post-Training Acceleration for Autoregressive Image Generation","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12617","citing_title":"SOAR: Self-Correction for Optimal Alignment and Refinement in Diffusion Models","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06966","citing_title":"MAR-GRPO: Stabilized GRPO for AR-diffusion Hybrid Image Generation","ref_index":31,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA","json":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA.json","graph_json":"https://pith.science/api/pith-number/JBUWOOUAVABPVIYMQYAFGN2BOA/graph.json","events_json":"https://pith.science/api/pith-number/JBUWOOUAVABPVIYMQYAFGN2BOA/events.json","paper":"https://pith.science/paper/JBUWOOUA"},"agent_actions":{"view_html":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA","download_json":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA.json","view_paper":"https://pith.science/paper/JBUWOOUA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.11455&json=true","fetch_graph":"https://pith.science/api/pith-number/JBUWOOUAVABPVIYMQYAFGN2BOA/graph.json","fetch_events":"https://pith.science/api/pith-number/JBUWOOUAVABPVIYMQYAFGN2BOA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA/action/storage_attestation","attest_author":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA/action/author_attestation","sign_citation":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA/action/citation_signature","submit_replication":"https://pith.science/pith/JBUWOOUAVABPVIYMQYAFGN2BOA/action/replication_record"}},"created_at":"2026-07-05T10:49:34.178768+00:00","updated_at":"2026-07-05T10:49:34.178768+00:00"}