{"as_of":"2026-08-14T14:09:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:3cad1b32748d39c0b7831027e1dbbffa6e8b5c3d4547ff42e75533ba3349b059","coverage":[{"denominator":50,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":50,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-06T14:46:50.653337Z","state":"measured"},{"denominator":50,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":50,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-14T06:32:32.682623+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/2507.17904/citation-record","integrity":"/paper/2507.17904/integrity","json":"/paper/2507.17904/citation-record.json","paper":"/paper/2507.17904"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.471978Z","title":"Gpt-4 Technical Report","venue":null,"work_id":"b65697c5-0502-454b-bf79-6c409177b803","year":2023},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.426519Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:66b03ced1c0543de404f3e243a98a395c5741e136eee636473789a58192df128","observation_id":"d9444d54-13cc-43bd-9f66-fd100b13ebda","resolution":{"observed_at":"2026-08-06T14:46:51.476589Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.457296Z","title":"Accordion: Adaptive Gradient Communication via Critical Learning Regime Identification","venue":null,"work_id":"16f59569-aa7c-4d7c-b4a0-63674ba590cd","year":2021},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.432286Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:9a3c4a1d3bb691bed5081ecb522c5f621afbf5bb2dded53e74a935b39e2b3e0c","observation_id":"1c41ff2f-4bc5-4352-963b-bbfe9164c39e","resolution":{"observed_at":"2026-08-06T14:46:51.462168Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.441380Z","title":"Sparse Communication for Distributed Gradient Descent","venue":null,"work_id":"1c4d3297-22d6-4a76-b83b-287a65b78efa","year":2017},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.437286Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:0a733f1efbf4bcf64bc038fee823d301c7f6675950eaa94b022d49ad6f3fb808","observation_id":"040664a4-1d3e-4029-8c3d-1673019026d8","resolution":{"observed_at":"2026-08-06T14:46:51.446605Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.427038Z","title":"QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding","venue":null,"work_id":"70f61c2b-fdca-4707-be81-d031dfe95f2b","year":2017},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.442234Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:3cab03a2cf39be4726b4f0d3e9463586b00437d33080d28f4c81487d7560711a","observation_id":"a0c7019c-1680-4f85-91d7-b823141e8800","resolution":{"observed_at":"2026-08-06T14:46:51.431758Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.410022Z","title":"The AI Disruption: Challenges and Guidance for Data Center Design","venue":null,"work_id":"7954aa54-2dda-49be-b1e4-b3f5390ecbdc","year":2023},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.447183Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:33481592eb18b2c7df1764be201fcd425c4bf2859fee9eda48639b2d7707ff0d","observation_id":"b0df2014-5acd-4035-b175-765ee9e4c0ac","resolution":{"observed_at":"2026-08-06T14:46:51.415162Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.394493Z","title":"Keyword Transformer: A Self-Attention Model for Keyword Spotting","venue":null,"work_id":"23a16a81-c5a7-4d25-9977-f06621cc165f","year":2021},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.452781Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:71324e1e4c9781f403ab51dab6b01edd7f2561961106048170552c809641bc58","observation_id":"62120cb1-655b-4527-a3f4-0b34709603ef","resolution":{"observed_at":"2026-08-06T14:46:51.399513Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.378883Z","title":"Flower: A Friendly Federated Learning Research Framework","venue":null,"work_id":"e2f5a40b-afdc-4af1-9e7d-a15346302d27","year":2020},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.458284Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:e7e1e4965606b1ae1b81e3a78bace42a497c31028f0ecdf50ed41ab0c57859bb","observation_id":"10538ecd-b018-4432-9f3a-74534ce9d53b","resolution":{"observed_at":"2026-08-06T14:46:51.383531Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.363290Z","title":"Towards Federated Learning at Scale: System Design","venue":null,"work_id":"a3bdbe1d-f03a-4b7c-8591-70b172e2fd2a","year":2019},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.462857Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:1ffc90a92f20848633761dc01e83816414c8b6594b5220f39d4030c4f536cf5c","observation_id":"828bc4ff-cd68-41b8-a4be-0698d04e8eee","resolution":{"observed_at":"2026-08-06T14:46:51.368348Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.348446Z","title":"Leaf: A Benchmark for Federated Settings","venue":null,"work_id":"39b1cd52-3607-482d-a572-ada0942f4cbf","year":2018},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.467397Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:38c23163254b50ff7724e25118578eb36917066573f981c787d096629590ce93","observation_id":"d54d3fd5-26e6-4bfc-b59b-e103ad9c3a62","resolution":{"observed_at":"2026-08-06T14:46:51.353242Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.333175Z","title":"FLAME: Federated Learning across Multi-device Environments","venue":null,"work_id":"ec184624-4a0e-4cd4-ad47-ce1f65241082","year":2022},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.471792Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:bb8103403f01a379093047093daea1bc0e69a4f51238313b325240d7d27ad51f","observation_id":"c6edfb8d-b72f-412b-8928-d36973caa7e9","resolution":{"observed_at":"2026-08-06T14:46:51.337799Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.317171Z","title":"EMNIST: Extending MNIST to Handwritten Letters","venue":null,"work_id":"9d2e2c6c-3481-43bf-95dc-0043feb27ed2","year":2017},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.476311Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:e744154922ebac91da72640dc88ef1a64b6b1e31417d981cc4e0d7d7d2730a43","observation_id":"8415f408-f770-4991-9e77-04dacf92a2a5","resolution":{"observed_at":"2026-08-06T14:46:51.322265Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.303178Z","title":"A Snapshot of the Frontiers of Client Selection in Federated Learning","venue":null,"work_id":"1a7f234c-eb97-4b3e-a20e-ee971f660bbf","year":2022},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.480682Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:344f4b284d945af4d93e301bed140651ba9f3e4934c8f1d755620e109a5d0bd3","observation_id":"d5484867-ed68-4c17-afe2-7f3bd9861f7a","resolution":{"observed_at":"2026-08-06T14:46:51.307575Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.289013Z","title":"Large Scale Distributed Deep Networks","venue":null,"work_id":"98ddceb7-c2f9-4388-ad03-2a49cab58f7f","year":2012},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.485077Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:40754f7000ca0baded6ac521e2c5a6985006ae03d6fdb47b9e412f0889dc9c5a","observation_id":"893df496-0414-4f08-8228-a68569c1f1ee","resolution":{"observed_at":"2026-08-06T14:46:51.293604Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.274385Z","title":"Exponential Laws of Computing Growth","venue":null,"work_id":"65a8a107-3a2a-4d16-9e48-5e7f7706f22e","year":2016},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.489584Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:676b9f43c4a616ff0f2973ea9d182491fad2a1034edbac1887fde0ca14a9bf42","observation_id":"144b08bb-767b-4018-983b-1aeb737d9485","resolution":{"observed_at":"2026-08-06T14:46:51.279068Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.259106Z","title":"QLoRA: Efficient Finetuning of Quantized LLMs","venue":null,"work_id":"79a7983a-1195-4b00-a934-331ab601afd3","year":2023},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.494160Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:ce383b4fb95791180a98fdfcfaeca3d812bd6c4f5b8d66b5cfe37348dd80ce1c","observation_id":"f83d6067-7a81-411b-93a0-6408522adf79","resolution":{"observed_at":"2026-08-06T14:46:51.263498Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.244438Z","title":"Beyond A Single AI Cluster: A Survey of Decentralized LLM Training","venue":null,"work_id":"0a91bd40-55b8-4adb-ba05-8d162fba01a3","year":2025},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.498780Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:99a2f67ddabfa0b8b36706abe2b5c858a5a831c45b34d3ab7f5a4f8aa913d588","observation_id":"39ff463c-611d-4929-b5f4-eb3d60eceb8f","resolution":{"observed_at":"2026-08-06T14:46:51.249293Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.229168Z","title":"DiLoCo: Distributed Low-Communication Training of Language Models","venue":null,"work_id":"bda82bc7-8e97-4160-b5fa-c5f8578919b9","year":2023},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.503182Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:015aa2322694971422a393b51aba3492cd96cd0cbcf0f1d93ca73b4428d511fb","observation_id":"09c160de-fe0d-4d2b-8a3e-30e8f6101c91","resolution":{"observed_at":"2026-08-06T14:46:51.234163Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.214554Z","title":"Client Selection in Federated Learning: Principles, Challenges, and Opportunities","venue":null,"work_id":"85f983fa-4276-4622-b0a5-4d992fee2c2c","year":2023},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.507661Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:d276294f0a9b34a62051033eb391676d0cbc988576d62c510a543b663930c0f0","observation_id":"63d7a928-e382-4513-8c8f-992d933ef017","resolution":{"observed_at":"2026-08-06T14:46:51.219626Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.198274Z","title":"TensorFlow Federated: Machine Learning on Decentralized Data","venue":null,"work_id":"fcafc2de-75dd-4057-905d-0680041859ac","year":2019},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.512122Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:9fd19e4db9d0802242a074eebafb855ba961d3cc29c364d72477f4d4ae3145c9","observation_id":"0630d78b-8c9a-4d32-9bfb-b6b13cadd60b","resolution":{"observed_at":"2026-08-06T14:46:51.203012Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.182955Z","title":"Efficiency: How We Do It","venue":null,"work_id":"0bd8aa4d-0848-41e8-80c9-1bdfffb2c183","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.516387Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:7131b919b1ade23ef03908ee02937360bb6f474ab068dc3f00e94a0e7423dba7","observation_id":"558270f2-167c-4498-8fda-cc3239886bf7","resolution":{"observed_at":"2026-08-06T14:46:51.188299Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.167270Z","title":"LoRA: Low-Rank Adaptation of Large Language Models","venue":null,"work_id":"8c39e555-1b3c-4e3c-8dd9-821d61b81836","year":2022},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.520845Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:217e1c155f2574c2c71367b1b23c2af0793378c5c39760e8793e5bdc2d76a703","observation_id":"da967d09-2c40-4731-98a7-6946aa041b5f","resolution":{"observed_at":"2026-08-06T14:46:51.172343Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.150034Z","title":"ThunderServe: High-Performance and Cost-Efficient LLM Serving in Cloud Environments","venue":null,"work_id":"57560131-f355-4ce6-8e53-9c541a64c22e","year":2025},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.525827Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:e00f285e898a227ee10d43c1ba2dca69aa6bfa01e4087a38e21f9db9637023b5","observation_id":"fe846704-0992-4c76-8378-50ffc5009935","resolution":{"observed_at":"2026-08-06T14:46:51.155589Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.133211Z","title":"MegaScale: Scaling Large Language Model Training to More than 10,000 GPUs","venue":null,"work_id":"56289b47-6932-4328-9353-70f2b9f26b16","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.530891Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:4509f0f2e439e41a708af265ede08a418fb2723c7178f4a388774e83305f4813","observation_id":"af67f297-beb7-4bbd-9a6b-a2591f033ea8","resolution":{"observed_at":"2026-08-06T14:46:51.138429Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.116382Z","title":"Scaling Laws for Neural Language Models","venue":null,"work_id":"699b0afc-affc-40c8-9463-da4afd9f7399","year":2020},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.535475Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:4ab5de7566bd6b5625abe6539cff281393253160214c68b021b204b4b997d3b1","observation_id":"7999ff26-fd8b-4a40-8a3d-95cd3981dbca","resolution":{"observed_at":"2026-08-06T14:46:51.121681Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.100193Z","title":"Federated Learning: Strategies for Improving Communication Efficiency","venue":null,"work_id":"46dc5e70-ff03-4005-8dde-2334523d3250","year":2016},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.539930Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:9befc6fe1aa9f1f6100f187634cd9149108f2f3b282a84a75f47ac2580e29cac","observation_id":"80c33391-2df9-4504-a4f9-6c5c6685980d","resolution":{"observed_at":"2026-08-06T14:46:51.105270Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.084191Z","title":"Oort: Efficient Federated Learning via Guided Participant Selection","venue":null,"work_id":"89501758-47fa-498d-a2c0-3aee80764b1b","year":2021},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.544503Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:22ec324722b071c5080f7765aefa475517ece90e73db6ff8e93b07c72771c04c","observation_id":"1e4eb50b-50b5-4b16-86fc-8fa4ba84b1f9","resolution":{"observed_at":"2026-08-06T14:46:51.089527Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.067752Z","title":"Breaking barriers to data center growth","venue":null,"work_id":"4d2778db-6be7-4794-9902-75544725ee8d","year":2025},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.549268Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:b24dd0da5a786e97e2d4dc857cd3ce14d83e9b6c674941f8fae96c6a1a27450e","observation_id":"ff0dc7b1-6bdc-400e-9a43-5e334c73cb2e","resolution":{"observed_at":"2026-08-06T14:46:51.072650Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.052132Z","title":"Federated Optimization in Heterogeneous Networks","venue":null,"work_id":"c2a84bde-56a9-4cc2-8aa0-c38cfc4052de","year":2020},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.554032Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:5438244fd8fd71ab28efcfe978f0a023bafaf7ef2c5fccbdc79155aa1c5ea16e","observation_id":"281ea7b2-3e2f-4189-a728-bd8fe03e85f9","resolution":{"observed_at":"2026-08-06T14:46:51.057348Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.036033Z","title":"Deep Gradient Compression: Reducing the Communication Bandwidth for Distributed Training","venue":null,"work_id":"e15e5243-3712-4c1c-969a-3ee457fb874d","year":2017},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.558487Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:e5127ab7a45bc5f7d326b65e47694806cfa04eba777007dba5faab4e67330254","observation_id":"ce04022f-f2de-461d-8bb7-68fa27b0051d","resolution":{"observed_at":"2026-08-06T14:46:51.041213Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.020781Z","title":"Ecolearn: Optimizing the Carbon Footprint of Federated Learning","venue":null,"work_id":"87e1aef5-23fd-4c32-be25-c54b2665224c","year":2023},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.563333Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:db0aebaa4f4feae80fd40058de697c51f463a754e0fbaa955783935dbf8c5feb","observation_id":"38965539-c5a1-4301-a92d-75187ec5ab81","resolution":{"observed_at":"2026-08-06T14:46:51.025542Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:51.004924Z","title":"Client Selection for Federated Learning with Heterogeneous Resources in Mobile Edge","venue":null,"work_id":"5e5557f5-a625-444b-906e-aa122ddd7c10","year":2019},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.567747Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:2da5af86bb074723c0316d259ad71534624724d69176bb89f329cf42dec8a1d1","observation_id":"bc0ab3c4-2094-470a-9b23-bcec9b71a068","resolution":{"observed_at":"2026-08-06T14:46:51.010395Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.989840Z","title":"Rethinking Load Growth: Assessing the Potential for Integration of Large Flexible Loads in US Power Systems","venue":null,"work_id":"f293c6d4-362b-45fe-99e6-f890f32ee93f","year":2025},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.571966Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:39ed80d5bd988d29b9187d4d824464341401eb31a32347a5d44ee8770b0d4bd6","observation_id":"9fd656cb-89a2-41dd-96e2-f8332be2ca2f","resolution":{"observed_at":"2026-08-06T14:46:50.994901Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.974255Z","title":"Carbon Emissions and Large Neural Network Training","venue":null,"work_id":"b9bcbd9b-b0d5-4fe1-b117-f6e10c6d63e0","year":2021},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.576337Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:1621fadd56137174653d1973c67b904fe2719e9227ffc58b996f1ae27b364876","observation_id":"6dcd0c25-9f0f-45c7-8aa5-10b2390b12c3","resolution":{"observed_at":"2026-08-06T14:46:50.979700Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.959328Z","title":"Expanding Data Center Capacity to Meet Growing Demand.McKinsey and Company, 2024","venue":null,"work_id":"6833594d-8911-4c89-9b60-21b618f27e60","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.580931Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:7599d143170ee0e7fb4b19c72cae61b3cda192f88a823cdc96b3b5637a244174","observation_id":"267e2e9a-b434-416e-b29f-9371c0b3a002","resolution":{"observed_at":"2026-08-06T14:46:50.964242Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.944170Z","title":"A Generic Framework for Privacy Preserving Deep Learning","venue":null,"work_id":"c21bd9ba-ed07-4ab1-b057-91c8ccab0b28","year":2018},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.585372Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:b2ecb98bb43268be1dddd9fb191af32c57b5036f7b18a714747f818a555981e3","observation_id":"1ae588f3-dfab-4257-9a40-52edbbe44175","resolution":{"observed_at":"2026-08-06T14:46:50.949053Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.928500Z","title":"AI, Data Centers and the Coming US Power De- mand Surge","venue":null,"work_id":"bc8efab7-a9cc-4f9e-b0a7-c660699e85c2","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.590157Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:0fe733ed98c3ba4b4132fdc1f05aee62f7de3617f4edef20158a6460ac0c777c","observation_id":"c04bc08b-64e8-4ad7-979a-83ab64ba1fbb","resolution":{"observed_at":"2026-08-06T14:46:50.933193Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.913513Z","title":"Data Center Power Oversubscription with a Medium Voltage Power Plane and Priority-Aware Capping","venue":null,"work_id":"0eb17c98-8445-453e-8042-e41dd0698497","year":2020},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.594977Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:ca10b8eb66e7a281f13fc2022f4d6ec7a90544fed8caa1aff2aaac4db80da71b","observation_id":"54cef338-5f3c-4ad3-965b-7fa948321ec7","resolution":{"observed_at":"2026-08-06T14:46:50.918309Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.897855Z","title":"Photon: Federated LLM Pre-Training","venue":null,"work_id":"b6a1eeac-509e-46d4-b644-c076250591e3","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.599927Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:afeae46e78b7a8072c71e698e3e2b2042cf8bcba7780032f16eec94e32f5abbe","observation_id":"7117c599-175e-4b2a-8f71-916db99d85b4","resolution":{"observed_at":"2026-08-06T14:46:50.902495Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.881518Z","title":"Green AI","venue":null,"work_id":"b0d552c4-cb1e-4713-bd1f-a6ac330112bd","year":2020},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.604234Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:562134a5ad958383017834b29770f3fe95e01120e29b7e2c3bc19e6cfe70a902","observation_id":"0da220d3-807c-4f27-908a-b6696627a1fd","resolution":{"observed_at":"2026-08-06T14:46:50.887858Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.865925Z","title":"A Quantitative Survey of Communication Optimizations in Distributed Deep Learning","venue":null,"work_id":"8a9ec96b-fd6f-4748-8383-496024402f13","year":2020},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.608798Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:936e3ed95f7dd6103bf9687416219973a8189b74353a61f0cddba9daadc3f612","observation_id":"0108eebd-a6d0-4676-9a60-ff5ebc0e8888","resolution":{"observed_at":"2026-08-06T14:46:50.871327Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.847582Z","title":"GreenLLM: Disaggregating Large Language Model Serving on Heterogeneous GPUs for Lower Carbon Emissions","venue":null,"work_id":"bdcfa108-9297-4c95-a2e2-b0bcd28aefd0","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.613030Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:ebd145aa7531984011f6c5140b3c7bed9fefb2fbcb570fb69d5df8f0929f725e","observation_id":"09aad996-cf2f-47e6-972d-653a6622a386","resolution":{"observed_at":"2026-08-06T14:46:50.853505Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.828060Z","title":"Democratizing AI: Open-Source Scalable LLM Training on GPU-Based Supercomputers","venue":null,"work_id":"2530a5e2-86eb-44a0-ab0b-570da9113be7","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.617626Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:f4890054f36397fd852f4c81d25f4da0e0f0129d9596720aab12801a35d84a5b","observation_id":"1fa96db6-5264-441b-9389-9f5748d57fd4","resolution":{"observed_at":"2026-08-06T14:46:50.834091Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.810226Z","title":"ML Training with Cloud GPU Shortages: Is Cross-Region the Answer? In MLSys, 2024","venue":null,"work_id":"db06e3ca-d809-4f77-a3af-3cb775a17404","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.622197Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:a8dbcbd9d16c508cfc90b380a6484d61e79cc3a0edaf79c6513785d95f0174c3","observation_id":"6cea097d-059c-4634-8e1f-a749208eea7f","resolution":{"observed_at":"2026-08-06T14:46:50.815463Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.791996Z","title":"Fusionllm: A Decentralized LLM Training System on Geo-Distributed GPUs with Adaptive Compression","venue":null,"work_id":"d506ef30-26f6-4f13-a058-eb594706bf16","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.626948Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:197dfd6003c284ab9167f8c3ee3ce2edbe0ddd8b2444049f81e02cd01ec75c00","observation_id":"fc46dfc4-3cb4-445e-ad47-44c9f1be9974","resolution":{"observed_at":"2026-08-06T14:46:50.797546Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.773174Z","title":"LlaMa 2: Open Foundation and Fine-Tuned Chat Models","venue":null,"work_id":"f446bf45-8736-4cc0-9c1a-ad2347fba6ed","year":2023},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.631325Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:bab0e1bb4e302041b9985addf1127bc17fbedbe50010eb53c6369c03cd001c51","observation_id":"ee66ed5d-7ed6-43dc-84c6-3dd3cf708d84","resolution":{"observed_at":"2026-08-06T14:46:50.779285Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.757078Z","title":"Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition","venue":null,"work_id":"4e819340-bbeb-4d8a-9be0-1753830fcd77","year":2018},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.636002Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:a970414f23090d3a2d12543941119c2ff80b4636766efb55e2090cbbb1989b64","observation_id":"21559192-b892-4476-bdf1-d3931fbf7a30","resolution":{"observed_at":"2026-08-06T14:46:50.762003Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.741205Z","title":"Sus- tainable AI: Environmental Implications, Challenges and Opportunities","venue":null,"work_id":"804dd77a-6613-4a74-a022-bcc4502ff347","year":2022},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.640265Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:ce5b3b1c4a3c415246fce28f5cca6306fa43defa5f6535bc96d42faf0eac7cbb","observation_id":"1409e08c-79f1-4fb2-a488-448351a653c6","resolution":{"observed_at":"2026-08-06T14:46:50.746577Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.723688Z","title":"SkyPilot: An Intercloud Broker for Sky Computing","venue":null,"work_id":"aa3739d3-d5a4-4d3d-b86d-f997fb3f2706","year":2023},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.644565Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:d24cad09d006483f2901727e1a289c8ae16e412965d44b7e39aac6d706ab0329","observation_id":"b1133fb2-4642-461c-9c3b-1a812004ecb5","resolution":{"observed_at":"2026-08-06T14:46:50.729341Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.705974Z","title":"Openfedllm: Training Large Language Models on Decentralized Private Data via Federated Learning","venue":null,"work_id":"1dfc1abe-992c-48e5-ba51-f10db7ec9ce1","year":2024},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.648918Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:77e752859fd747b2740852a6ae01004652eb55e979109c310f09d2a5607c28ef","observation_id":"d7cedff9-b7c9-4daf-b066-9c866d88ab6a","resolution":{"observed_at":"2026-08-06T14:46:50.711468Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T14:46:50.685856Z","title":"OPT: Open Pre-Trained Transformer Language Models","venue":null,"work_id":"4d52b69b-eb81-4644-9a20-0f242d635999","year":2022},"citing_paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-06T14:46:50.653337Z"},"links":{"citing_paper":"/paper/2507.17904"},"observation_digest":"sha256:5bae07a37c3489fb43b2d5f475899bb228bb98a6fd17993e2cfac57c172d820c","observation_id":"d972b638-49f8-4ce0-a530-0214fd934132","resolution":{"observed_at":"2026-08-06T14:46:50.694184Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2507.17904","last_updated":"2025-07-23T20:02:06Z","latest_version":1,"primary_category":"cs.DC","snapshot_observed_at":"2026-08-07T11:03:04.692969Z","submitted_at":"2025-07-23T20:02:06Z","title":"PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training"},"reference_resolution":{"displayed":50,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":0,"verified_exact":0,"verified_fuzzy":50},"total_outbound_references":50},"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":"2026-08-14T06:32:32.682623+00:00","source":"crossref"},{"observed_at":"2026-08-14T06:32:18.44784+00:00","source":"retraction_watch"}],"thesis":"As of 14 August 2026, this Paper Citation Record lists 50 of 50 outbound references and 0 inbound Pith citation observations for arXiv:2507.17904."}