{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UJTTIYFBGOTOLORU42RPT63MA4","short_pith_number":"pith:UJTTIYFB","schema_version":"1.0","canonical_sha256":"a2673460a133a6e5ba34e6a2f9fb6c073db1386ec89c9537188a36a3d04981c3","source":{"kind":"arxiv","id":"2506.08210","version":1},"attestation_state":"computed","paper":{"title":"A Comprehensive Study of Decoder-Only LLMs for Text-to-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Andrew Z. Wang, Ming-Yu Liu, Songwei Ge, Tero Karras, Yogesh Balaji","submitted_at":"2025-06-09T20:29:53Z","abstract_excerpt":"Both text-to-image generation and large language models (LLMs) have made significant advancements. However, many text-to-image models still employ the somewhat outdated T5 and CLIP as their text encoders. In this work, we investigate the effectiveness of using modern decoder-only LLMs as text encoders for text-to-image diffusion models. We build a standardized training and evaluation pipeline that allows us to isolate and evaluate the effect of different text embeddings. We train a total of 27 text-to-image models with 12 different text encoders to analyze the critical aspects of LLMs that cou"},"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":"2506.08210","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-09T20:29:53Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"158d023b40e8ba071be122afd415809f65eabb7432d3ca1f713ce4ed405bb3ec","abstract_canon_sha256":"ebb0609e16fa5ac1886bb8bbaf629918be4e1c4f8a4d33134bb65219c0e80a69"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:21:48.819711Z","signature_b64":"2JFAOtqR1EIduQ2d6f63BoPl/IK1kP0ZzEpO3EI9sPsZ6QTRMcKUFon7rMrdK9F9P51+AAjAB+S78Z6JL7FXAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a2673460a133a6e5ba34e6a2f9fb6c073db1386ec89c9537188a36a3d04981c3","last_reissued_at":"2026-07-05T11:21:48.819133Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:21:48.819133Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Comprehensive Study of Decoder-Only LLMs for Text-to-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Andrew Z. Wang, Ming-Yu Liu, Songwei Ge, Tero Karras, Yogesh Balaji","submitted_at":"2025-06-09T20:29:53Z","abstract_excerpt":"Both text-to-image generation and large language models (LLMs) have made significant advancements. However, many text-to-image models still employ the somewhat outdated T5 and CLIP as their text encoders. In this work, we investigate the effectiveness of using modern decoder-only LLMs as text encoders for text-to-image diffusion models. We build a standardized training and evaluation pipeline that allows us to isolate and evaluate the effect of different text embeddings. We train a total of 27 text-to-image models with 12 different text encoders to analyze the critical aspects of LLMs that cou"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.08210","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/2506.08210/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":"2506.08210","created_at":"2026-07-05T11:21:48.819196+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.08210v1","created_at":"2026-07-05T11:21:48.819196+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.08210","created_at":"2026-07-05T11:21:48.819196+00:00"},{"alias_kind":"pith_short_12","alias_value":"UJTTIYFBGOTO","created_at":"2026-07-05T11:21:48.819196+00:00"},{"alias_kind":"pith_short_16","alias_value":"UJTTIYFBGOTOLORU","created_at":"2026-07-05T11:21:48.819196+00:00"},{"alias_kind":"pith_short_8","alias_value":"UJTTIYFB","created_at":"2026-07-05T11:21:48.819196+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/UJTTIYFBGOTOLORU42RPT63MA4","json":"https://pith.science/pith/UJTTIYFBGOTOLORU42RPT63MA4.json","graph_json":"https://pith.science/api/pith-number/UJTTIYFBGOTOLORU42RPT63MA4/graph.json","events_json":"https://pith.science/api/pith-number/UJTTIYFBGOTOLORU42RPT63MA4/events.json","paper":"https://pith.science/paper/UJTTIYFB"},"agent_actions":{"view_html":"https://pith.science/pith/UJTTIYFBGOTOLORU42RPT63MA4","download_json":"https://pith.science/pith/UJTTIYFBGOTOLORU42RPT63MA4.json","view_paper":"https://pith.science/paper/UJTTIYFB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.08210&json=true","fetch_graph":"https://pith.science/api/pith-number/UJTTIYFBGOTOLORU42RPT63MA4/graph.json","fetch_events":"https://pith.science/api/pith-number/UJTTIYFBGOTOLORU42RPT63MA4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UJTTIYFBGOTOLORU42RPT63MA4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UJTTIYFBGOTOLORU42RPT63MA4/action/storage_attestation","attest_author":"https://pith.science/pith/UJTTIYFBGOTOLORU42RPT63MA4/action/author_attestation","sign_citation":"https://pith.science/pith/UJTTIYFBGOTOLORU42RPT63MA4/action/citation_signature","submit_replication":"https://pith.science/pith/UJTTIYFBGOTOLORU42RPT63MA4/action/replication_record"}},"created_at":"2026-07-05T11:21:48.819196+00:00","updated_at":"2026-07-05T11:21:48.819196+00:00"}