{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:TTWPTCLF73VTHWTHTU7DKTBP72","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"005f4259b6fe21a22155a1478a2114325e997777214ca29617d05d8e5a8e6fad","cross_cats_sorted":["cs.CV","cs.CY"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-21T08:12:47Z","title_canon_sha256":"27cd6c96af1f6e27af6aacab836b4eb0447ee92fc058e070a7a2efd42f56eed7"},"schema_version":"1.0","source":{"id":"2511.17031","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2511.17031","created_at":"2026-05-18T03:10:11Z"},{"alias_kind":"arxiv_version","alias_value":"2511.17031v2","created_at":"2026-05-18T03:10:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.17031","created_at":"2026-05-18T03:10:11Z"},{"alias_kind":"pith_short_12","alias_value":"TTWPTCLF73VT","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"TTWPTCLF73VTHWTH","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"TTWPTCLF","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:60a3d02ab15e206ef0109b3b36438a8bd2230383b8b9161b2a7c9e3edfb4203c","target":"graph","created_at":"2026-05-18T03:10:11Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Our energy scaling law achieves high predictive accuracy within individual architectures (R² > 0.9) and exhibits strong cross-architecture generalization, maintaining high rank correlations across models and enabling reliable energy estimation for unseen model–hardware combinations."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"denoising operations dominate energy consumption due to their repeated execution across multiple inference steps"},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"An adapted scaling law predicts GPU energy consumption for diffusion model inference with R² > 0.9 within architectures and strong cross-architecture generalization."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"An adapted Kaplan scaling law predicts GPU energy use for diffusion models from FLOPs."}],"snapshot_sha256":"f1610e83350e27929b9bf8f24125ca5ab371c356a9a98065bac18dfad14f256b"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"728af761ae654365c46a9a5c61163214dda24cae6a1ac1b74daaa2e0483a77e7"},"paper":{"abstract_excerpt":"The rapidly growing computational demands of diffusion models for image generation have raised significant concerns about energy consumption and environmental impact. While existing approaches to energy optimization focus on architectural improvements or hardware acceleration, there is a lack of principled methods to predict energy consumption across different model configurations and hardware setups. We propose an adaptation of Kaplan scaling laws to predict GPU energy consumption for diffusion models based on computational complexity (FLOPs). Our approach decomposes diffusion model inference","authors_text":"Aniketh Iyengar, Boris Ruf, Jiaqi Han, Marcin Detyniecki, Stefano Ermon, Vincent Grari","cross_cats":["cs.CV","cs.CY"],"headline":"An adapted Kaplan scaling law predicts GPU energy use for diffusion models from FLOPs.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-21T08:12:47Z","title":"Energy Scaling Laws for Diffusion Models: Quantifying Compute in Image Generation"},"references":{"count":33,"internal_anchors":8,"resolved_work":33,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"The Gentle Singularity","work_id":"1e63cffa-0dd7-42bc-8c4b-1e677cbfa6d8","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Wolff Anthony, Benjamin Kanding, and Raghavendra Selvan","work_id":"5266ac43-fd62-4de0-beb0-39a11e93a12f","year":2007},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"de Araújo, JPW, and MinervaBooks","work_id":"0174ac4e-54eb-462b-881a-848755f10d8c","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Measuring the environmental impact of delivering AI at google scale","work_id":"1266f22b-4371-4adc-a0da-6cd5ff7625c5","year":2025},{"cited_arxiv_id":"2403.03206","doi":"","is_internal_anchor":true,"ref_index":5,"title":"Scaling Rectified Flow Transformers for High-Resolution Image Synthesis","work_id":"4dc55d76-271e-42dd-878f-c20546599c69","year":2024}],"snapshot_sha256":"60516dc159bb1e3ce6d0947ecf139eaa2cccb71419f3eec3038f4a064bf39008"},"source":{"id":"2511.17031","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-17T21:03:21.452941Z","id":"4764e91c-955c-4d1a-a785-fe01c184cd8b","model_set":{"reader":"grok-4.3"},"one_line_summary":"An adapted scaling law predicts GPU energy consumption for diffusion model inference with R² > 0.9 within architectures and strong cross-architecture generalization.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"An adapted Kaplan scaling law predicts GPU energy use for diffusion models from FLOPs.","strongest_claim":"Our energy scaling law achieves high predictive accuracy within individual architectures (R² > 0.9) and exhibits strong cross-architecture generalization, maintaining high rank correlations across models and enabling reliable energy estimation for unseen model–hardware combinations.","weakest_assumption":"denoising operations dominate energy consumption due to their repeated execution across multiple inference steps"}},"verdict_id":"4764e91c-955c-4d1a-a785-fe01c184cd8b"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:bf4d36bcd040d9b861a29fdde5b9df36d1e128a63db503151fd1ad3854ba7fc2","target":"record","created_at":"2026-05-18T03:10:11Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"005f4259b6fe21a22155a1478a2114325e997777214ca29617d05d8e5a8e6fad","cross_cats_sorted":["cs.CV","cs.CY"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-21T08:12:47Z","title_canon_sha256":"27cd6c96af1f6e27af6aacab836b4eb0447ee92fc058e070a7a2efd42f56eed7"},"schema_version":"1.0","source":{"id":"2511.17031","kind":"arxiv","version":2}},"canonical_sha256":"9cecf98965feeb33da679d3e354c2ffe82dcd5e5e297158998c01681294b0200","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9cecf98965feeb33da679d3e354c2ffe82dcd5e5e297158998c01681294b0200","first_computed_at":"2026-05-18T03:10:11.796250Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:10:11.796250Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"k2X4cPT+7PhOEvc5918weNjPX0SYc+HXykV6Bj/aOfbcWEsAEZGIaJLsCpZdrlr4Wwzirnj3FAVe0/GCddGnAA==","signature_status":"signed_v1","signed_at":"2026-05-18T03:10:11.797031Z","signed_message":"canonical_sha256_bytes"},"source_id":"2511.17031","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bf4d36bcd040d9b861a29fdde5b9df36d1e128a63db503151fd1ad3854ba7fc2","sha256:60a3d02ab15e206ef0109b3b36438a8bd2230383b8b9161b2a7c9e3edfb4203c"],"state_sha256":"07f4f657756f713bf5c9a58df545a66f1e8aa20520493a13ad63cd60460ccd8f"}