{"paper":{"title":"Shap-E: Generating Conditional 3D Implicit Functions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Shap-E generates parameters of implicit functions for 3D assets directly from text prompts.","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Alex Nichol, Heewoo Jun","submitted_at":"2023-05-03T23:59:13Z","abstract_excerpt":"We present Shap-E, a conditional generative model for 3D assets. Unlike recent work on 3D generative models which produce a single output representation, Shap-E directly generates the parameters of implicit functions that can be rendered as both textured meshes and neural radiance fields. We train Shap-E in two stages: first, we train an encoder that deterministically maps 3D assets into the parameters of an implicit function; second, we train a conditional diffusion model on outputs of the encoder. When trained on a large dataset of paired 3D and text data, our resulting models are capable of"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"When trained on a large dataset of paired 3D and text data, our resulting models are capable of generating complex and diverse 3D assets in a matter of seconds.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The encoder stage can deterministically map arbitrary 3D assets into implicit-function parameters with negligible information loss, allowing the diffusion stage to operate in a faithful latent space.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Shap-E encodes 3D assets into implicit function parameters then uses a conditional diffusion model to generate new ones from text, enabling fast multi-representation 3D asset creation.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Shap-E generates parameters of implicit functions for 3D assets directly from text prompts.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"be0b99523b105351dba2745c1f12536c0934e252b31f1227cd477a22df1e1437"},"source":{"id":"2305.02463","kind":"arxiv","version":1},"verdict":{"id":"b8eca8b0-c0fd-4c1a-bd8d-9db32b723a91","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T15:27:46.007085Z","strongest_claim":"When trained on a large dataset of paired 3D and text data, our resulting models are capable of generating complex and diverse 3D assets in a matter of seconds.","one_line_summary":"Shap-E encodes 3D assets into implicit function parameters then uses a conditional diffusion model to generate new ones from text, enabling fast multi-representation 3D asset creation.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The encoder stage can deterministically map arbitrary 3D assets into implicit-function parameters with negligible information loss, allowing the diffusion stage to operate in a faithful latent space.","pith_extraction_headline":"Shap-E generates parameters of implicit functions for 3D assets directly from text prompts."},"references":{"count":74,"sample":[{"doi":"","year":2017,"title":"Learning Representations and Generative Models for 3D Point Clouds","work_id":"cfca29dd-dc5d-468c-82c9-b34d1626c543","ref_index":1,"cited_arxiv_id":"1707.02392","is_internal_anchor":true},{"doi":"","year":2023,"title":"MusicLM: Generating Music From Text","work_id":"15e6566e-1c36-468f-966e-823248cbf87f","ref_index":2,"cited_arxiv_id":"2301.11325","is_internal_anchor":true},{"doi":"","year":2023,"title":"Sine: Semantic-driven image-based nerf editing with prior-guided editing field","work_id":"dc9dac60-a1c1-4d48-8c09-84a79db31c24","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Gaudi: A neural architect for immersive 3d scene generation","work_id":"c2453470-f4b6-48c9-96b8-0c6867b32eb0","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Audiolm: a language modeling approach to audio generation","work_id":"3a2aea1c-8117-4436-9100-0536aea3cf47","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":74,"snapshot_sha256":"5d8ef6dbaf41d7fa61aeb0216a3adb0eae85fd1761e379fbea19bf2e532ebaa1","internal_anchors":33},"formal_canon":{"evidence_count":2,"snapshot_sha256":"c55b0d76488cb727a41297b1494d15c8a3b79caf50ff097a3dbf931da3ac4b3e"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}