{"as_of":"2026-08-23T01:49:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:65f8dea8f45e95be13f180f880e493d921cf2b3468a9e6d0cce45653e3dc57de","coverage":[{"denominator":53,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":53,"source":"paper_references, paper_reference_links","source_observed_at":"2026-06-28T18:23:25.570021Z","state":"measured"},{"denominator":53,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":53,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-22T06:32:14.747728+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2606.00567/citation-record","integrity":"/paper/2606.00567/integrity","json":"/paper/2606.00567/citation-record.json","paper":"/paper/2606.00567"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"A scalable processing- in-memory accelerator for parallel graph processing,","venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:709cd6756cb047027881f3e4358e7fbbc54eb8d5880da022343c390cedec2a58","observation_id":"da9257bd-9d87-4c20-8cd0-f035cbce492b","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Stable video diffusion: Scaling latent video diffusion models to large datasets,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:f0d168c1318d18e2d245784684a8b2fade6239a5d34946720d1ef37122f8bdc4","observation_id":"3ef307d4-8e8f-4f32-84f6-e4f6a575fe91","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Token merging for fast stable diffusion,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:7a5980bdabe77a6fdb05ef221f96ef8f3164c01277da26f0b5933c13706a7fff","observation_id":"5c092a73-4117-4eff-b490-3c679901c189","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2401.09047","last_updated":"2024-01-17T08:30:32Z","snapshot_observed_at":"2026-08-21T22:46:57.008517Z","submitted_at":"2024-01-17T08:30:32Z","title":"VideoCrafter2: Overcoming Data Limitations for High-Quality Video Diffusion Models","version":1},"cited_work":{"arxiv_id":"2401.09047","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2401.09047","snapshot_observed_at":"2026-06-28T20:42:37.758214Z","title":"Videocrafter2: Overcoming data limita- tions for high-quality video diffusion models","venue":null,"work_id":"8c2c1d92-75e1-4574-929f-379bc5317044","year":2024},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"cited_paper":"/paper/2401.09047","citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:f9fd21e7ff54afddb1326d6ce2e94b1a292bacd5d4ec5f18e841182aab1c734b","observation_id":"aa4dba0c-0520-4ab7-9817-6dea9b083c72","resolution":{"observed_at":"2026-06-28T20:42:37.759728Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"WaveGrad: Estimating gradients for waveform generation,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:a02ca2ae69d3f90388e24d61ad0f80ba41115767cf975e6aa9431d1ccbcbb04d","observation_id":"015b087a-9fba-47fc-96ca-64b25637d294","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"DianNao: A small-footprint high-throughput accelerator for ubiqui- tous machine-learning,","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:66aa0fbb02125cff521330612949cd4ba6c6efc15a25af788aa5c4b6477f0240","observation_id":"0c3a13a5-9705-4201-9437-c08b17d27cd4","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Executing your commands via motion diffusion in latent space,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:ddedac61bcf83ab5a171c38fcb2a0d6af5929094cd1cac262f7001c619de7f67","observation_id":"5f0f7d2b-fd2f-407a-954b-5f87adc98f10","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Eyeriss: A spatial architecture for energy-efficient dataflow for convolutional neural networks,","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:36f5e1dfd048fd696f054fe156c74ab77a7f4ea296a1b651dc948f0e55da1efc","observation_id":"2e823ab0-ca06-4ade-b7c0-c9a71eeadfa6","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"DaDianNao: A machine-learning supercomputer,","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:7c81ff7d0c588d39df0cb68c57f429dc502dc14d39915e408519ccca98ae0587","observation_id":"e9083f7d-b4c2-4360-9886-a48d1d954591","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"LLMServingSim: A HW/SW co-simulation infrastructure for LLM inference serving at scale,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:b565400c90b034fe9a89489a702431722ee80e5544841c4f5f6d52b836d2b02c","observation_id":"42a4f6d4-6d1a-45a8-abfe-a6e8bdfdd37e","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Diffusion models beat GANs on image synthesis,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:de2e362c06e7010ae5167c365f9954709d323bfd6300018c47e933773a24ebcf","observation_id":"bd3dea1e-fcc7-4c02-b2e4-8c7c0d3703bd","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Graphi- cionado: A high-performance and energy-efficient accelerator for graph analytics,","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:299a3d797c31b28743987fc5f534cd6b58fb1ef9c534b71f8d9de9e1d2a7d1b5","observation_id":"faa11666-bf8a-4861-9b83-1b095e7ae036","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"EIE: Efficient inference engine on compressed deep neural network,","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:e6049da5cc33d38ed45a16863fe78c2b287e34d08320cc0927ad4048f7b11ec0","observation_id":"fa26f506-54b2-44ff-ba8e-68e91d477683","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"ExTensor: An accelerator for sparse tensor algebra,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:c639ae2890f8ad79bad948cccc46b87a7f1af0cf39aea8e7f8c5b632b5b49f97","observation_id":"c93d1891-95fe-48de-ad03-6ed5b0b108a7","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"EXION: Exploiting inter- and intra-iteration output sparsity for diffusion models,","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:9663f16f3012249defe6d3152ed384c51ca73f049d1dc9f41b9d10392180433f","observation_id":"219b7578-81e0-42d7-8241-8c4bd43a8bd7","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Denoising diffusion probabilistic models,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:f8b64aeea76daaebf786aaa48b1288030f0a4efcadb9dd74958b7cb933853abf","observation_id":"84eb1ff9-d0dd-4f42-9322-19c0c6ad34a1","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Video diffusion models,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:a79824cefddaa02e21d7d867cd68e83c19f642b1c187b7d81975c2db333e941e","observation_id":"5b0c97af-3a5e-4bed-90cd-b25c7d5ff011","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2301.12661","last_updated":"2023-01-30T04:44:34Z","snapshot_observed_at":"2026-08-17T08:10:22.173010Z","submitted_at":"2023-01-30T04:44:34Z","title":"Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models","version":1},"cited_work":{"arxiv_id":"2301.12661","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2301.12661","snapshot_observed_at":"2026-06-28T20:42:37.763677Z","title":"Make-An-Audio: Text-to-audio generation with prompt-enhanced diffusion models","venue":null,"work_id":"a9bc5106-f11c-415f-9d34-fb7a465c5185","year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"cited_paper":"/paper/2301.12661","citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:1bd241ed5cdfdca66e05107140cd79b5ef8eec8efddc5c6dafb008ec6f94f62a","observation_id":"21a39f3c-96e9-43ae-b160-44cbb19326f1","resolution":{"observed_at":"2026-06-28T20:42:37.765250Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"In-datacenter performance analysis of a tensor processing unit,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:edd9834f4028210b38d249b235b022f7b5051b8c1e044a25c51ca7fa67a939e1","observation_id":"71756340-1269-4395-bfb5-451f1c62854d","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Elucidating the design space of diffusion-based generative models,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:349df1780c31222f405d98c7740fa71ae3df68b9dbf32f25e37b8a337cea0064","observation_id":"ac0e1dc9-2763-4ded-9d92-02d377f517e0","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Ramulator: A fast and extensible DRAM simulator,","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:958bdc8db3843dbd60c1fc18d93e31751f8ff3881badeb2c23200ceb2f2aa261","observation_id":"2048ba63-1a08-450f-9b6c-2c3ad26a349e","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Cambricon-D: Full-network differential acceleration for diffusion models,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:21863e1365cf4e611e87728a4105fabda462bdf742d6dc8f6dc82ee4e037c2a4","observation_id":"22e57950-4fec-4fa6-8a22-54aa9a0c9012","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"DiffWave: A versatile diffusion model for audio synthesis,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:6c282c4dc873dc091850f892e93826607c9706440298f79d21c38dccf3db010e","observation_id":"48b2c09a-0d55-42cb-af7c-c4d61f0e3d27","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"MAERI: Enabling flexible dataflow mapping over DNN accelerators via reconfigurable intercon- nects,","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:5c202ffacc699fd68a3883d2ee32994afd5a337f5083ee743911a6e27e3629aa","observation_id":"7677c806-4593-4395-bee8-fe4d13685b0c","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Q-Diffusion: Quantizing diffusion models,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:edb7de8c0795487c63211edda1faf96566fae0e22b24c450ba365f2186d12bdb","observation_id":"75df79f6-b9df-4ac8-ba51-e683d7ca798d","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"SnapFusion: Text-to-image diffusion model on mobile devices within two seconds,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:ab748bc48a09559985648683e820187d8a2d8a187bc1c382c15f062368e2581d","observation_id":"f168f077-7304-4c17-ada4-2e1e55f14599","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Pseudo numerical methods for diffusion models on manifolds,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:368dd381aceac2ad7d1d495a31ac55336dd93c59891b1b9d26cdcd3ae4e01622","observation_id":"0f1b85a4-1e87-4170-8cf2-a2221371fc69","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2211.01095","last_updated":"2025-05-19T07:56:56Z","snapshot_observed_at":"2026-08-21T07:38:18.450127Z","submitted_at":"2022-11-02T13:14:30Z","title":"DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models","version":3},"cited_work":{"arxiv_id":"2211.01095","doi":null,"metadata_source":"pith","pith_arxiv_id":"2211.01095","snapshot_observed_at":"2026-07-11T01:57:50.075044Z","title":"DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models","venue":"cs.LG","work_id":"a7e16614-fbfc-45d6-8fb6-714035cb707c","year":2022},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"cited_paper":"/paper/2211.01095","citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:16525313a24a5c5caa8b3a3bb8f848b60ed5ed09f0d7d4564215aed6654f2eee","observation_id":"76771843-40d8-498c-8d96-7c52f2277e77","resolution":{"observed_at":"2026-06-28T20:42:37.762264Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Ramulator 2.0: A modern, modular, and extensible DRAM simulator,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:5b7c8215c741296580d19c688b2f7d6d342a55cdb0823f997a3ef0827821213d","observation_id":"fe2a8848-8c31-48ef-aa6a-08b2654dac93","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"DeepCache: Accelerating diffusion models for free,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:7b369f1fc3084d43d1aa28be87a5f07dc3f3679b7232ad3e8319ff2ecc5e2fda","observation_id":"26ff7ec5-7bb0-4cf5-9598-28e0b6c982c7","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Improved denoising diffusion probabilistic models,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:050611d4702841cff450f977bfcbf1148871860a27e92e215647005842077a0b","observation_id":"98c6faf9-a80c-4ff9-8979-e01aeeab239e","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"OuterSPACE: An outer product based sparse matrix multiplication accelerator,","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:c861be2e912828febb76d945ab20c99362fd1b05900ce82e54e8cf0465c1abf8","observation_id":"f082c3c1-2932-4ae5-b9cd-f61d61dfc097","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Timeloop: A systematic approach to DNN accelerator evaluation,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:8baa9e211cfdfbbfe28fa165256a2e41d62d05a3beaeb589b8d988e33e8812d7","observation_id":"639d7b87-3228-4bbb-9892-0f6740cd2438","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"SCNN: An accelerator for compressed-sparse convolutional neural networks,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:63a88f6cef8fca30652bcb516b95b9c9bf62039f17a0dc0e3ef3c394d1ebe92e","observation_id":"fbfa65a4-412f-424c-abb7-5d35aa91362d","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Tale of two Cs: Computation vs. communication scaling for future trans- formers on future hardware,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:8411aacc8ca590945ecd46d58c16d379d49d0a1b3dad4845094acd7dcf9c4bab","observation_id":"0506e373-9f2d-45e7-91b5-f740e94700b7","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Scalable diffusion models with transformers,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:9fdbb70244b614d6962282e7c600bfcddc4408a4c7c9fe115c558438e7cd17ab","observation_id":"dbecae90-90db-4729-acee-2b33c0a1ee99","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"SIGMA: A sparse and irregular GEMM accelerator with flexible interconnects for DNN training,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:185fd6e765f7e64969c8aa576410ed0001b69e30e0fe937e07cc0010545a5c82","observation_id":"0d00b54c-ba29-4a9e-a692-e5b7cf05b93b","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Hierar- chical text-conditional image generation with CLIP latents,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:664c9e468c363b2e9a56ae38eb5f1493e212846a96ae3f037a94bba16d832606","observation_id":"18742bd9-2e6d-47d9-9195-a5e2a662f5ae","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"MLPerf inference benchmark,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:7b756e5446c4b214fed5d3143f76848a6aa78b94f8fb09772fa59cd19f8c1568","observation_id":"16d21aac-80be-4355-b9f7-330329f7ac00","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"High-resolution image synthesis with latent diffusion models,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:b54416e012dc7d10d8431a6d8a591c9e5fcf4f7fda83fe2ee1b931c7a3f7a44b","observation_id":"1f34ac8d-1331-43f3-af6b-c4cb1d598ab5","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"DRAMSim2: A cycle accurate memory system simulator,","venue":null,"work_id":null,"year":2011},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:80f7f120d79f33e0e581780061e7f78666fd4f961d395c48279d735e1bb50b73","observation_id":"2ecae5bb-0921-4b97-9c3f-aa30a1cc4d9a","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Photorealistic text-to-image diffusion models with deep language understanding,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:618502c7019c25beb914e7c6724e0934754270bc51f4c54d49d51df9d23ee6df","observation_id":"0411b9c0-065a-4225-aac5-c1b23ab96274","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Progressive distillation for fast sampling of diffusion models,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:104a7f2347d3e0b5189634b50081807b03ecf8e23988bb80bb7e292776849324","observation_id":"74a5576a-2595-464e-a157-4ea0b4b725f5","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Post-training quantization on diffusion models,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:ee911a3ffdc32a5f214959108d38fa9023962c2ff0cbc0eb3ddac3f1ed29d774","observation_id":"f03c872e-1aae-4231-805d-a9f7add11115","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Denoising diffusion implicit models,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:be9b520ce1029ac23c3ee41a9d72a7101ccb64f74350f6f596df772b8e956cc0","observation_id":"04b74385-f78d-4f7b-81f1-94118afeab15","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Score-based generative modeling through stochastic differential equations,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:80adda222d1285f174107438c0a7e2b4f54e26405ec21032c72ee5eb67a4202b","observation_id":"bb24d47e-5cb7-4196-95e9-5c578c85d6f7","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"MatRaptor: A sparse-sparse matrix multiplication accelerator based on row- wise product,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:98458682bcf94f7904cb995ff610f85d06f9fd89ec90f25bf0857e11fc06e742","observation_id":"c386f1c0-222f-4be4-9ecc-68561b46627b","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Human motion diffusion model,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:fff789905550d233a050cbf42682dd60e9b9372847bab4d204117a6c0ae8f606","observation_id":"f89d696f-87e6-4627-84f6-08ab26210c61","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"EDGE: Editable dance generation from music,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:8190a32560bd8f73bc5196d4fbffddabfbcadbfb824283a311598cb9758a65da","observation_id":"b7acc231-f687-4ab8-8f3d-f4a8266964cb","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Trapezoid: A versatile accelerator for dense and sparse matrix multiplications,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:1213ef606a1d31bec082c5104580361a2b8e00f1ea24fad9af132110b5cd7738","observation_id":"48b311ef-f301-46f7-9e6e-898f77a54f18","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Gamma: Leveraging gustavson’s algorithm to accelerate sparse matrix multiplication,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:ae3bb152b402e8b398ce7d5310e5e2d7fcb33c4848a1b821bf33ae18c728e251","observation_id":"396fdda9-fe62-4290-9eb6-ca08f5ed392e","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"Cambricon-X: An accelerator for sparse neural networks,","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:9ce4019a93f4401144ec069027b4071a80e89126967904ee4ed08e9a4f183cd2","observation_id":"c268d5d3-06b0-4f80-b64f-1fad4c19b1c3","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-06-28T18:23:25.570021Z","title":"SpArch: Efficient architecture for sparse matrix multiplication,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures","version":2},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-06-28T18:23:25.570021Z"},"links":{"citing_paper":"/paper/2606.00567"},"observation_digest":"sha256:f23f08deeb4eacf7154029453fbd8e8ce172518d796a292d8c453d6d1c2e8bb3","observation_id":"1089e0cf-52e1-453e-a291-3db2ceb35cea","resolution":{"observed_at":"2026-06-28T18:23:25.570021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2606.00567","last_updated":"2026-06-02T04:39:39Z","latest_version":2,"primary_category":"cs.AR","snapshot_observed_at":"2026-08-18T04:08:40.981852Z","submitted_at":"2026-05-30T06:38:54Z","title":"Activation Concentration: Characterizing Column-Level Output Sparsity Across Diffusion Model Architectures"},"reference_resolution":{"displayed":53,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":50,"verified_exact":3,"verified_fuzzy":0},"total_outbound_references":53},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"thesis":"As of 23 August 2026, this Paper Citation Record lists 53 of 53 outbound references and 0 inbound Pith citation observations for arXiv:2606.00567."}