{"as_of":"2026-08-10T13:17:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:d105bad9a7e31b22e0ac4a4de1aabcf2b6e2dba3a8ac9d5a3d282bc620f5e60b","coverage":[{"denominator":30,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":30,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-04T23:34:24.887183Z","state":"measured"},{"denominator":30,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":30,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-10T06:31:04.303077+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/2509.06535/citation-record","integrity":"/paper/2509.06535/integrity","json":"/paper/2509.06535/citation-record.json","paper":"/paper/2509.06535"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-04T23:34:22.716638Z","title":"Optuna: A next-generation hyperparameter optimization framework","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":1,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:22.716638Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:a0931edbd0a48dc2373a67029873c3f3fd8188067a9b2b27d39784a79486fd96","observation_id":"9b4f8c76-1ea0-4eaf-b763-99e86445916c","resolution":{"observed_at":"2026-08-04T23:34:22.716638Z","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":"10.18653/v1/2022.aacl-main.61","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"A prompt array keeps the bias away: Debiasing vision-language models with adversarial learning","venue":null,"work_id":"0f76614c-7509-4e1e-941d-f68d6eeb9c89","year":2022},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":2,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:22.779424Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:7fbb3a08357db3b214def96b973c7a569ed26dfb8ea1516e1e7a30bc98cabe63","observation_id":"34791a7e-a35d-4be0-970e-b32abf8e6dd6","resolution":{"observed_at":"2026-08-04T23:34:28.171113Z","resolver_source":"doi","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1707.00075","last_updated":"2017-07-07T01:31:36Z","snapshot_observed_at":"2026-07-06T05:49:19.425378Z","submitted_at":"2017-07-01T01:09:33Z","title":"Data Decisions and Theoretical Implications when Adversarially Learning Fair Representations","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1707.00075","snapshot_observed_at":"2026-08-04T23:34:22.839522Z","title":null,"venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":3,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:22.839522Z"},"links":{"cited_paper":"/paper/1707.00075","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:dd8c7802abdcec9027f3e1a445a10f456d3550b5c312705f5cc9d1f063213a35","observation_id":"3865cc76-2607-4386-ba3c-94736cea9c25","resolution":{"observed_at":"2026-08-04T23:34:22.839522Z","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":"2026-08-04T23:34:22.918504Z","title":"Fairness in machine learning: A survey","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":4,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:22.918504Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:0a5faba22ced15d44c4d58d41573f28beb7a5910bfe3b3aba63ebbb52c5830de","observation_id":"545e49b6-2b0a-4b7f-b8ca-d43573b4d86c","resolution":{"observed_at":"2026-08-04T23:34:22.918504Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-04T23:34:27.810926Z","title":"Sinkhorn distances: Lightspeed computation of optimal transport","venue":null,"work_id":"93409379-d566-4502-b9c0-c39e0b0b1065","year":2013},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":5,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:22.967496Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:57f80b0152505a31ea43c101efb0669e007af3b8a5de92c8996408b8888ce4cc","observation_id":"5f33a531-874c-499f-8670-2ab66f0b8014","resolution":{"observed_at":"2026-08-04T23:34:27.922478Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:27.592897Z","title":"Interpolating between optimal transport and mmd using sinkhorn divergences","venue":null,"work_id":"91eaf11d-f978-4df7-bc47-61baece32078","year":2019},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":6,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.042064Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:b794975e5c9cec2d1fa0cd20e36a6d67a5e316d268b252ace063b408d5f14328","observation_id":"70eceb34-c108-461c-9c63-5b76ebc29eaf","resolution":{"observed_at":"2026-08-04T23:34:27.688235Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:27.381325Z","title":"Examining gender and racial bias in large vision -- language models using a novel dataset of parallel images","venue":null,"work_id":"094e3f87-0e1b-4d23-a2ad-1dc2c8b3720c","year":2024},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":7,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.090240Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:f426d370034cdf1e2e2f96e67b5fd66bb3ae7ad658665f0bd1fe5345455d4f1e","observation_id":"5fad9c70-be37-48e6-bfa9-339a3319f93d","resolution":{"observed_at":"2026-08-04T23:34:27.479997Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":"2021.10037","doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-04T23:34:25.754269Z","title":"Role of machine learning in medical research: A survey","venue":null,"work_id":"35301fd7-5a19-46cd-b739-92e66d92770c","year":2021},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":8,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.171366Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:6b0f85596a58181fa4ef8509ae59acd9f7ff72b40fd846089d210b0d5f7f6cd8","observation_id":"0a6fe612-95ab-4c79-9779-56d40c5f28e8","resolution":{"observed_at":"2026-08-04T23:34:25.835930Z","resolver_source":"raw_fallback","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:27.173250Z","title":"Borgwardt, Malte J","venue":null,"work_id":"dd6abef0-bf5e-4602-82fb-46e21ce8f0e4","year":2012},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":9,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.232245Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:fa5fe06be7b613cfb6c6dbd0e175adc5dde3ef17a99ef6d5af3af22db9e41c77","observation_id":"a2cd5f86-8782-44cb-a723-d27fa0e0ca80","resolution":{"observed_at":"2026-08-04T23:34:27.270742Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:26.967259Z","title":"Duurzaam in de e-infrastructuur","venue":null,"work_id":"72998590-324a-4d57-bf48-d3847878fcc9","year":2017},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":10,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.281371Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:f329d97b39b160a414ab253c62e2e1f74aabf115e1214ac132b40dc71b86d568","observation_id":"26e9e90a-e99f-4a2d-9050-d873e243cbef","resolution":{"observed_at":"2026-08-04T23:34:27.060494Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:23.484282Z","title":"Batch normalization: accelerating deep network training by reducing internal covariate shift","venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":13,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.484282Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:5eb8865a3c1aa05b0e9031a66ed17c901e6e756b68df1bbc417ead1c4eddf8dd","observation_id":"16586fa4-b667-4f73-a74e-87aadc64f2ed","resolution":{"observed_at":"2026-08-04T23:34:23.484282Z","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":"2026-08-04T23:34:23.551349Z","title":"Fairness-aware classifier with prejudice remover regularizer","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":14,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.551349Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:145030190af5da32c4e4a137ba4b72e8e730ce513a95145cf95ea839cacbb9b3","observation_id":"36cdc8c3-2d85-4731-a633-7a3f507689ce","resolution":{"observed_at":"2026-08-04T23:34:23.551349Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1908.04913","last_updated":"2019-08-14T01:42:41Z","snapshot_observed_at":"2026-08-09T16:05:36.101423Z","submitted_at":"2019-08-14T01:42:41Z","title":"FairFace: Face Attribute Dataset for Balanced Race, Gender, and Age","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1908.04913","snapshot_observed_at":"2026-08-04T23:34:23.621390Z","title":"Fairface: Face attribute dataset for balanced race, gender, and age for bias measurement and mitigation","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":15,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.621390Z"},"links":{"cited_paper":"/paper/1908.04913","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:fe59b917253040ffa57b8285cea5725ef1ad76ad65ae4440e392004c2a4c16b8","observation_id":"26291ade-e883-4449-9c95-98506e219ab4","resolution":{"observed_at":"2026-08-04T23:34:23.621390Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1412.6980","last_updated":"2017-01-30T01:27:54Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2014-12-22T13:54:29Z","title":"Adam: A Method for Stochastic Optimization","version":9},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1412.6980","snapshot_observed_at":"2026-08-04T23:34:23.700398Z","title":"Kingma and Jimmy Ba","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":16,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.700398Z"},"links":{"cited_paper":"/paper/1412.6980","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:4e253e9a7a7dc4b36bef57fc8d1d016a94183c8ba966f792b1cd247bfc02804e","observation_id":"470e6c4c-c799-46bd-82cc-6283d92eab0a","resolution":{"observed_at":"2026-08-04T23:34:23.700398Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1910.09700","last_updated":"2019-11-04T20:37:33Z","snapshot_observed_at":"2026-07-06T08:31:12.309726Z","submitted_at":"2019-10-21T23:57:32Z","title":"Quantifying the Carbon Emissions of Machine Learning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1910.09700","snapshot_observed_at":"2026-08-04T23:34:23.741510Z","title":"Quantifying the carbon emissions of machine learning","venue":null,"work_id":null,"year":1910},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":17,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.741510Z"},"links":{"cited_paper":"/paper/1910.09700","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:392459b82f6997af96bc20201871cc191db584989adf83d716f738b636ca6df7","observation_id":"caa55965-674f-4281-b39e-b6f4654e0b65","resolution":{"observed_at":"2026-08-04T23:34:23.741510Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2309.14381","last_updated":"2023-09-24T15:34:56Z","snapshot_observed_at":"2026-08-05T04:28:43.652726Z","submitted_at":"2023-09-24T15:34:56Z","title":"Survey of Social Bias in Vision-Language Models","version":1},"cited_work":{"arxiv_id":"2309.14381","doi":null,"metadata_source":"pith","pith_arxiv_id":"2309.14381","snapshot_observed_at":"2026-08-04T23:34:25.385600Z","title":"Survey of Social Bias in Vision-Language Models","venue":"cs.CL","work_id":"08b7a4de-acaa-4977-8573-0b39fb90b6bd","year":2023},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":18,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.816344Z"},"links":{"cited_paper":"/paper/2309.14381","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:076b16f7fcfe3df521bc16b90d96bab32a463d4f6a6af1caf7e8aa105740693b","observation_id":"eeffc51e-89c3-46f6-8990-10d43954e3c7","resolution":{"observed_at":"2026-08-04T23:34:25.442697Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:26.713989Z","title":null,"venue":null,"work_id":"e7d0b345-b2a2-440c-bf1f-5d5c9329f6a8","year":2023},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":19,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.878176Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:027b415e9020627dba3f50fe35097c863dd87c2b451a38de6991e7e1151c4520","observation_id":"e363e6ca-3157-4de3-b3f3-a9b7b77bac8d","resolution":{"observed_at":"2026-08-04T23:34:26.818862Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2301.12597","last_updated":"2023-06-15T07:57:29Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-01-30T00:56:51Z","title":"BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2301.12597","snapshot_observed_at":"2026-08-04T23:34:23.952939Z","title":"BLIP-2: bootstrapping language-image pre-training with frozen image encoders and large language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":20,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:23.952939Z"},"links":{"cited_paper":"/paper/2301.12597","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:03568f8178ca908a0eda1b10bc61a137526c0cd4bea4f3bc8c7925bdd6186f82","observation_id":"ded0135d-d234-456c-b198-69c59b631eac","resolution":{"observed_at":"2026-08-04T23:34:23.952939Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2403.19949","last_updated":"2024-04-05T20:08:16Z","snapshot_observed_at":"2026-07-06T17:52:54.875134Z","submitted_at":"2024-03-29T03:15:31Z","title":"FairCLIP: Harnessing Fairness in Vision-Language Learning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.19949","snapshot_observed_at":"2026-08-04T23:34:24.026585Z","title":"Fairclip: Harnessing fairness in vision-language learning","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":21,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.026585Z"},"links":{"cited_paper":"/paper/2403.19949","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:8aed1c7d756892ec5d9bb1d05fc927bd9587e9b5a7e8f7ecf506a52bc3fa9ef7","observation_id":"90c65bd5-2c56-4382-8b90-1eeda03b7903","resolution":{"observed_at":"2026-08-04T23:34:24.026585Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-04T23:34:26.481308Z","title":"Methodiek co2 emissiefactoren elektriciteit","venue":null,"work_id":"400d16cb-fb52-4444-8eb8-a373d2fd2a5c","year":2024},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":22,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.103981Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:107126f11f8d0c041dc4fcb0a13f69617cdb6a7796c0f2c5c02f3a60dfbea616","observation_id":"b7d44bc9-fc22-48a5-bae5-7941db218019","resolution":{"observed_at":"2026-08-04T23:34:26.593891Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2303.08774","last_updated":"2024-03-04T06:01:33Z","snapshot_observed_at":"2026-08-07T07:30:12.213965Z","submitted_at":"2023-03-15T17:15:04Z","title":"GPT-4 Technical Report","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.08774","snapshot_observed_at":"2026-08-04T23:34:24.150894Z","title":"Gpt-4 technical report, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":23,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.150894Z"},"links":{"cited_paper":"/paper/2303.08774","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:1b7ad1d5a07c4990866caa9ce8f588aa8470bba8a45b912200f1fe285efe0191","observation_id":"4112c3fc-42b9-4524-a14b-e1f811eee073","resolution":{"observed_at":"2026-08-04T23:34:24.150894Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1910.11779","last_updated":"2019-10-25T15:03:11Z","snapshot_observed_at":"2026-07-06T08:32:17.643291Z","submitted_at":"2019-10-25T15:03:11Z","title":"Toward a better trade-off between performance and fairness with kernel-based distribution matching","version":1},"cited_work":{"arxiv_id":"1910.11779","doi":null,"metadata_source":"pith","pith_arxiv_id":"1910.11779","snapshot_observed_at":"2026-08-04T23:34:25.241689Z","title":"Toward a better trade-off between performance and fairness with kernel-based distribution matching","venue":"cs.LG","work_id":"9c0dc7fb-c0e4-4bf0-9383-508d5250fa45","year":2019},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":24,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.225324Z"},"links":{"cited_paper":"/paper/1910.11779","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:2b0df782a54351b9ee3bb0abf2e33141260c49c7aa0cdbba0c6d7d8d3555f38e","observation_id":"c5c6e83f-ce6e-4dd2-8d01-c440d5a524d8","resolution":{"observed_at":"2026-08-04T23:34:25.296677Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:26.264112Z","title":"Recycling privileged learning and distribution matching for fairness","venue":null,"work_id":"deaaefa7-66a4-4099-a7f8-517af477a886","year":2017},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":25,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.302581Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:5a2dab61bf52d57fd062d06923a4428976439340274fb536b91b061f34166e21","observation_id":"90ff2445-3556-4208-b5d3-a3ffb012ed9e","resolution":{"observed_at":"2026-08-04T23:34:26.376918Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:26.003314Z","title":"Learning transferable visual models from natural language supervision","venue":null,"work_id":"69c8dd8e-8299-41d1-8a39-5828ade3c351","year":2021},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":26,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.378415Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:4d3d7f3eeb0e624828ee93fe7e680cbe9a92fd8f86f9a16efca478266f622fd0","observation_id":"8963bc3a-9300-4614-a76c-7ce4456ad8da","resolution":{"observed_at":"2026-08-04T23:34:26.118360Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2303.10431","last_updated":"2023-03-18T14:57:43Z","snapshot_observed_at":"2026-07-06T15:05:12.470716Z","submitted_at":"2023-03-18T14:57:43Z","title":"DeAR: Debiasing Vision-Language Models with Additive Residuals","version":1},"cited_work":{"arxiv_id":"2303.10431","doi":"10.48550/arxiv.2303.10431","metadata_source":"pith","pith_arxiv_id":"2303.10431","snapshot_observed_at":"2026-08-05T02:49:54.815029Z","title":"DeAR: Debiasing Vision-Language Models with Additive Residuals","venue":"cs.CV","work_id":"8e963000-9365-41ae-8922-d5371b196653","year":2023},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":27,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.476189Z"},"links":{"cited_paper":"/paper/2303.10431","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:779ab0a0bec46f9e62e3e3c96278c432b25ea59d2f0924801023e18753c53243","observation_id":"8c392530-8a2e-4cbd-9821-fad59e433aa7","resolution":{"observed_at":"2026-08-04T23:34:25.080998Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+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":"2026-08-04T23:34:24.550895Z","title":"Representation bias in data: A survey on identification and resolution techniques","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":28,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.550895Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:c86a46c0a99ea90880da2de47b53e64554f9e0b636b559a2eea0979aff3e0195","observation_id":"6b393817-6e9b-40a6-8012-2b7391a5bcdb","resolution":{"observed_at":"2026-08-04T23:34:24.550895Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1708.03888","last_updated":"2017-09-13T23:25:07Z","snapshot_observed_at":"2026-07-06T05:55:05.946111Z","submitted_at":"2017-08-13T11:01:57Z","title":"Large Batch Training of Convolutional Networks","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1708.03888","snapshot_observed_at":"2026-08-04T23:34:24.631571Z","title":"Large batch training of convolutional networks, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":29,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.631571Z"},"links":{"cited_paper":"/paper/1708.03888","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:cb6bccc094b6e6842ac4dbd507561769080ff26727bad93dbffd8ef29eeca332","observation_id":"6b69d579-96dc-4258-9a3f-02eab965d4c2","resolution":{"observed_at":"2026-08-04T23:34:24.631571Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2312.07353","last_updated":"2025-03-26T13:32:28Z","snapshot_observed_at":"2026-07-31T04:25:37.437501Z","submitted_at":"2023-12-12T15:21:57Z","title":"CLIP in Medical Imaging: A Survey","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2312.07353","snapshot_observed_at":"2026-08-04T23:34:24.706167Z","title":"Clip in medical imaging: A comprehensive survey, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":30,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.706167Z"},"links":{"cited_paper":"/paper/2312.07353","citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:a08d50923d1aef60b2373f86e5d0b3203675e7ac6899981cfe0a84927cc3201d","observation_id":"dc49f2ab-4eb6-4fde-8a61-3be8845ad56d","resolution":{"observed_at":"2026-08-04T23:34:24.706167Z","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":"2026-08-04T23:34:24.820294Z","title":"General facial representation learning in a visual-linguistic manner","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":31,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.820294Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:1be778b60c3e745317b01d230489820fb4d662ef5c07fcbf384177560fb041f9","observation_id":"e9dbd783-2b62-44d4-9647-eab22e467e91","resolution":{"observed_at":"2026-08-04T23:34:24.820294Z","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":"2026-08-04T23:34:24.887183Z","title":"write newline","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''","version":1},"reference_index":32,"source":"arxiv_source","source_observed_at":"2026-08-04T23:34:24.887183Z"},"links":{"citing_paper":"/paper/2509.06535"},"observation_digest":"sha256:cac1eeb4789a965cb784b6b58014a910db46272af37960a308b25ce33ef05f61","observation_id":"64fcf9ef-90f0-4acf-96da-67616df9b647","resolution":{"observed_at":"2026-08-04T23:34:24.887183Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2509.06535","last_updated":"2025-09-08T10:41:10Z","latest_version":1,"primary_category":"cs.CV","snapshot_observed_at":"2026-08-09T16:06:26.457762Z","submitted_at":"2025-09-08T10:41:10Z","title":"On the Reproducibility of \"FairCLIP: Harnessing Fairness in Vision-Language Learning''"},"reference_resolution":{"displayed":30,"state_counts":{"malformed_identifier":0,"metadata_mismatch":2,"parse_uncertain":0,"unresolved":17,"verified_exact":3,"verified_fuzzy":8},"total_outbound_references":30},"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":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"thesis":"As of 10 August 2026, this Paper Citation Record lists 30 of 30 outbound references and 0 inbound Pith citation observations for arXiv:2509.06535."}