{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:XK2B4L2IJZSROB2JWL2GOTJYIE","short_pith_number":"pith:XK2B4L2I","schema_version":"1.0","canonical_sha256":"bab41e2f484e65170749b2f4674d38410d8598493107df9a674d431824565de5","source":{"kind":"arxiv","id":"2606.00583","version":1},"attestation_state":"computed","paper":{"title":"Improving Visual Representation Alignment Generation with GRPO","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MM"],"primary_cat":"cs.CV","authors_text":"Shentong Mo, Sukmin Yun","submitted_at":"2026-05-30T07:21:40Z","abstract_excerpt":"Recent diffusion transformers have demonstrated strong image synthesis capabilities but remain inefficient to train due to weak alignment between generative and discriminative representations. While representation alignment frameworks such as REPA improve convergence by aligning noisy denoising features with pretrained visual encoders, their externally supervised alignment loss is static and lacks adaptivity during training and inference. Existing methods rely on fixed cosine alignment or contrastive objectives, which cannot dynamically balance representation consistency and generation quality"},"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":"2606.00583","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-30T07:21:40Z","cross_cats_sorted":["cs.AI","cs.LG","cs.MM"],"title_canon_sha256":"41b1d66c4608abbb422ca282952502e4b31741d692a2e9dd59a19e84536806af","abstract_canon_sha256":"e951c7f9b87e071cb829005c564eb89ac9bfe4ad47f3af403e77b36573c7a324"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:58.919362Z","signature_b64":"lCMF1niSGy/+S+n15GoPTMQ70DwfFVT6SNVJiuZFtdW2+eVxV05pL5V5YBn2FBbaljvtpayyJ4x2wWK2hFJlBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bab41e2f484e65170749b2f4674d38410d8598493107df9a674d431824565de5","last_reissued_at":"2026-06-02T01:03:58.918982Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:58.918982Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Visual Representation Alignment Generation with GRPO","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MM"],"primary_cat":"cs.CV","authors_text":"Shentong Mo, Sukmin Yun","submitted_at":"2026-05-30T07:21:40Z","abstract_excerpt":"Recent diffusion transformers have demonstrated strong image synthesis capabilities but remain inefficient to train due to weak alignment between generative and discriminative representations. While representation alignment frameworks such as REPA improve convergence by aligning noisy denoising features with pretrained visual encoders, their externally supervised alignment loss is static and lacks adaptivity during training and inference. Existing methods rely on fixed cosine alignment or contrastive objectives, which cannot dynamically balance representation consistency and generation quality"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.00583","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/2606.00583/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":"2606.00583","created_at":"2026-06-02T01:03:58.919028+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.00583v1","created_at":"2026-06-02T01:03:58.919028+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.00583","created_at":"2026-06-02T01:03:58.919028+00:00"},{"alias_kind":"pith_short_12","alias_value":"XK2B4L2IJZSR","created_at":"2026-06-02T01:03:58.919028+00:00"},{"alias_kind":"pith_short_16","alias_value":"XK2B4L2IJZSROB2J","created_at":"2026-06-02T01:03:58.919028+00:00"},{"alias_kind":"pith_short_8","alias_value":"XK2B4L2I","created_at":"2026-06-02T01:03:58.919028+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/XK2B4L2IJZSROB2JWL2GOTJYIE","json":"https://pith.science/pith/XK2B4L2IJZSROB2JWL2GOTJYIE.json","graph_json":"https://pith.science/api/pith-number/XK2B4L2IJZSROB2JWL2GOTJYIE/graph.json","events_json":"https://pith.science/api/pith-number/XK2B4L2IJZSROB2JWL2GOTJYIE/events.json","paper":"https://pith.science/paper/XK2B4L2I"},"agent_actions":{"view_html":"https://pith.science/pith/XK2B4L2IJZSROB2JWL2GOTJYIE","download_json":"https://pith.science/pith/XK2B4L2IJZSROB2JWL2GOTJYIE.json","view_paper":"https://pith.science/paper/XK2B4L2I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.00583&json=true","fetch_graph":"https://pith.science/api/pith-number/XK2B4L2IJZSROB2JWL2GOTJYIE/graph.json","fetch_events":"https://pith.science/api/pith-number/XK2B4L2IJZSROB2JWL2GOTJYIE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XK2B4L2IJZSROB2JWL2GOTJYIE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XK2B4L2IJZSROB2JWL2GOTJYIE/action/storage_attestation","attest_author":"https://pith.science/pith/XK2B4L2IJZSROB2JWL2GOTJYIE/action/author_attestation","sign_citation":"https://pith.science/pith/XK2B4L2IJZSROB2JWL2GOTJYIE/action/citation_signature","submit_replication":"https://pith.science/pith/XK2B4L2IJZSROB2JWL2GOTJYIE/action/replication_record"}},"created_at":"2026-06-02T01:03:58.919028+00:00","updated_at":"2026-06-02T01:03:58.919028+00:00"}