{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:ZITBY3TXX5KVHFYEWVDXCCOT3Q","short_pith_number":"pith:ZITBY3TX","schema_version":"1.0","canonical_sha256":"ca261c6e77bf55539704b5477109d3dc3b019b7aaeedf6549cdefcb5dc833ad6","source":{"kind":"arxiv","id":"2210.10960","version":2},"attestation_state":"computed","paper":{"title":"Diffusion Models already have a Semantic Latent Space","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jaeseok Jeong, Mingi Kwon, Youngjung Uh","submitted_at":"2022-10-20T02:07:23Z","abstract_excerpt":"Diffusion models achieve outstanding generative performance in various domains. Despite their great success, they lack semantic latent space which is essential for controlling the generative process. To address the problem, we propose asymmetric reverse process (Asyrp) which discovers the semantic latent space in frozen pretrained diffusion models. Our semantic latent space, named h-space, has nice properties for accommodating semantic image manipulation: homogeneity, linearity, robustness, and consistency across timesteps. In addition, we introduce a principled design of the generative proces"},"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":"2210.10960","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2022-10-20T02:07:23Z","cross_cats_sorted":[],"title_canon_sha256":"66ad54f58a329254187f3586fe9a1eeaa14293aa7bffbbae4222f172282a84e3","abstract_canon_sha256":"5e5e7cebd25456beb6787d44eacb4b4c472173715ef95dcafb491fe17f9d4682"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:55:59.918175Z","signature_b64":"JjPG7aDKk/Ytm577IMYUjQZQIf3c0QyW+aG4bGVM5jPrpW4wGr35fA0tyOyQBUzEdv151JHvWfLabsD98o+YDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ca261c6e77bf55539704b5477109d3dc3b019b7aaeedf6549cdefcb5dc833ad6","last_reissued_at":"2026-07-05T05:55:59.917803Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:55:59.917803Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Diffusion Models already have a Semantic Latent Space","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jaeseok Jeong, Mingi Kwon, Youngjung Uh","submitted_at":"2022-10-20T02:07:23Z","abstract_excerpt":"Diffusion models achieve outstanding generative performance in various domains. Despite their great success, they lack semantic latent space which is essential for controlling the generative process. To address the problem, we propose asymmetric reverse process (Asyrp) which discovers the semantic latent space in frozen pretrained diffusion models. Our semantic latent space, named h-space, has nice properties for accommodating semantic image manipulation: homogeneity, linearity, robustness, and consistency across timesteps. In addition, we introduce a principled design of the generative proces"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.10960","kind":"arxiv","version":2},"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/2210.10960/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":"2210.10960","created_at":"2026-07-05T05:55:59.917859+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.10960v2","created_at":"2026-07-05T05:55:59.917859+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.10960","created_at":"2026-07-05T05:55:59.917859+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZITBY3TXX5KV","created_at":"2026-07-05T05:55:59.917859+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZITBY3TXX5KVHFYE","created_at":"2026-07-05T05:55:59.917859+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZITBY3TX","created_at":"2026-07-05T05:55:59.917859+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08337","citing_title":"AutoAnchor: Stable Diffusion Unlearning Using Cross-Attention as a Manifold Surrogate","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2606.27371","citing_title":"Don't Settle at the Mode! Mitigating Diversity Collapse in Pretrained Flow Models via Feature Self-Guidance","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13240","citing_title":"Towards More General Control of Diffusion Models Using Jeffrey Guidance","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31053","citing_title":"AnchorSteer: Self-Discovered Concept Injection for Structure-Preserving Music Editing","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27343","citing_title":"Towards Controllable Image Generation through Representation-Conditioned Diffusion Models","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18150","citing_title":"Whispers in the Noise: Surrogate-Guided Concept Awakening via a Multi-Agent Framework","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19511","citing_title":"Are Watermarked Images Editable? SafeMark for Watermark-Preserving Text-Guided Image Editing","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2509.18831","citing_title":"Text Slider: Efficient and Plug-and-Play Continuous Concept Control for Image/Video Synthesis via LoRA Adapters","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10439","citing_title":"Filtering Memorization from Parameter-Space in Diffusion Models","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25119","citing_title":"Evaluation without Generation: Non-Generative Assessment of Harmful Model Specialization with Applications to CSAM","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06367","citing_title":"The Interplay of Data Structure and Imbalance in the Learning Dynamics of Diffusion Models","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18804","citing_title":"Geometric Decoupling: Diagnosing the Structural Instability of Latent","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17673","citing_title":"Grokking of Diffusion Models: Case Study on Modular Addition","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q","json":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q.json","graph_json":"https://pith.science/api/pith-number/ZITBY3TXX5KVHFYEWVDXCCOT3Q/graph.json","events_json":"https://pith.science/api/pith-number/ZITBY3TXX5KVHFYEWVDXCCOT3Q/events.json","paper":"https://pith.science/paper/ZITBY3TX"},"agent_actions":{"view_html":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q","download_json":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q.json","view_paper":"https://pith.science/paper/ZITBY3TX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.10960&json=true","fetch_graph":"https://pith.science/api/pith-number/ZITBY3TXX5KVHFYEWVDXCCOT3Q/graph.json","fetch_events":"https://pith.science/api/pith-number/ZITBY3TXX5KVHFYEWVDXCCOT3Q/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q/action/storage_attestation","attest_author":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q/action/author_attestation","sign_citation":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q/action/citation_signature","submit_replication":"https://pith.science/pith/ZITBY3TXX5KVHFYEWVDXCCOT3Q/action/replication_record"}},"created_at":"2026-07-05T05:55:59.917859+00:00","updated_at":"2026-07-05T05:55:59.917859+00:00"}