{"id":"7a84a12fb5bfd31974109de52d19c361efc8f23695076ff401d43497cd638a91","title":"Infrathon: The authors test language agents on long-horizon AI infrastructure optimization","argument":"The authors of Infrathon ask whether frontier agents can keep improving real-world AI infrastructure over extended work periods. Their benchmark covers diverse AI infrastructure optimization tasks, and they evaluate six models across 150 runs with 24-hour budgets. They report that model rankings vary with the evaluation horizon, and that 135 runs improved further in the second half of their budgets. In their results, shorter evaluations can miss both later gains and shifts in which model performs best. The authors also report a case where Claude Opus 5.5 raised the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%. The comparison suggests agents may find infrastructure solutions that outperform human-designed baselines on a measured task; it does not establish that result across other tasks or deployments. The authors identify task violations, unreliable testing and premature completion as barriers to sustained improvement. The abstract presents Infrathon as a resource for evaluating agents over longer periods and discovering infrastructure solutions.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The supplied evidence is the paper's abstract, which does not detail the benchmark tasks, evaluation methods, or how the reported barriers affected individual runs. The reported codec comparison is for one stated task; the abstract does not establish performance across other tasks or real-world deployments.","speaker":{"name":"Hu, Zhengyu, Xiao, Zheyuan, Wang, Yike, Ivison, Hamish, Wang, Haoyu, Yuan, Yige, Shao, Rulin, Li, Yuetai, Lyu, Niqi, Ding, Yicheng, Yang, Shu, Zeng, Zhichen, Zhu, Banghua, Pei, Jiaxin, Ding, Kaize, Dasigi, Pradeep, Roth, Dan, Hajishirzi, Hannaneh, Smith, Noah A., Xiao, Teng, Poovendran, Radha","roleAtSource":"Article author"},"source":{"title":"Infrathon: Benchmarking Language Agents for Real-World, Long-Horizon AI Infra Optimization","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.11995","publishedAt":"2026-11-09","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors introduce Infrathon to study whether frontier agents can continually improve real-world AI infrastructure over long horizons.","excerpt":"In this work, we study whether frontier agent can continually improve real-world AI infrastructure over long horizons. We introduce Infrathon","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"The authors evaluate six models in 150 runs, each with a 24-hour budget, totaling over 10,000 GPU-hours.","excerpt":"We evaluate six models across 150 runs with 24-hour budgets, totaling over 10,000 GPU-hours.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"Model rankings vary with evaluation horizon, and 135 of 150 runs improved further during the second half of their budgets.","excerpt":"model rankings vary with the evaluation horizon, while 135 of the 150 runs achieve further improvements during the second half of their budgets.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"The authors say shorter evaluations can miss both optimization gains and changes in relative model performance.","excerpt":"These findings indicate that shorter evaluations can overlook both optimization gains and changes in relative model performance.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"The abstract reports Claude Opus 5.5 increased mean KV-cache compression ratio from 6.82% with a human-designed codec to 19.86%.","excerpt":"Claude Opus 5.5 increases the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"The authors identify task violations, unreliable testing and premature completion as barriers to sustained improvement.","excerpt":"Analysis of agent trajectories identifies task violations, unreliable testing, and premature completion as barriers to sustained improvement.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.11995","checkedAt":"2026-11-12T10:50:08.144Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How much does each of task violations, unreliable testing and premature completion limit sustained improvement in the benchmark runs?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"Infrathon: The authors test language agents on long-horizon AI infrastructure optimization","summary":"The authors of Infrathon ask whether frontier agents can keep improving real-world AI infrastructure over extended work periods. Their benchmark covers diverse AI infrastructure optimization tasks, and they evaluate six models across 150 runs with 24-hour budgets. They report that model rankings vary with the evaluation horizon, and that 135 runs improved further in the second half of their budgets. In their results, shorter evaluations can miss both later gains and shifts in which model performs best. The authors also report a case where Claude Opus 5.5 raised the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%. The comparison suggests agents may find infrastructure solutions that outperform human-designed baselines on a measured task; it does not establish that result across other tasks or deployments. The authors identify task violations, unreliable testing and premature completion as barriers to sustained improvement. The abstract presents Infrathon as a resource for evaluating agents over longer periods and discovering infrastructure solutions.","question":"How much does each of task violations, unreliable testing and premature completion limit sustained improvement in the benchmark runs?","claims":[{"text":"The authors introduce Infrathon to study whether frontier agents can continually improve real-world AI infrastructure over long horizons.","excerpt":"In this work, we study whether frontier agent can continually improve real-world AI infrastructure over long horizons. We introduce Infrathon","support":"source-stated"},{"text":"The authors evaluate six models in 150 runs, each with a 24-hour budget, totaling over 10,000 GPU-hours.","excerpt":"We evaluate six models across 150 runs with 24-hour budgets, totaling over 10,000 GPU-hours.","support":"source-stated"},{"text":"Model rankings vary with evaluation horizon, and 135 of 150 runs improved further during the second half of their budgets.","excerpt":"model rankings vary with the evaluation horizon, while 135 of the 150 runs achieve further improvements during the second half of their budgets.","support":"source-stated"},{"text":"The authors say shorter evaluations can miss both optimization gains and changes in relative model performance.","excerpt":"These findings indicate that shorter evaluations can overlook both optimization gains and changes in relative model performance.","support":"source-stated"},{"text":"The abstract reports Claude Opus 5.5 increased mean KV-cache compression ratio from 6.82% with a human-designed codec to 19.86%.","excerpt":"Claude Opus 5.5 increases the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%.","support":"source-stated"},{"text":"The authors identify task violations, unreliable testing and premature completion as barriers to sustained improvement.","excerpt":"Analysis of agent trajectories identifies task violations, unreliable testing, and premature completion as barriers to sustained improvement.","support":"source-stated"}],"limitations":["The supplied evidence is the paper's abstract, which does not detail the benchmark tasks, evaluation methods, or how the reported barriers affected individual runs.","The reported codec comparison is for one stated task; the abstract does not establish performance across other tasks or real-world deployments."],"mediaId":"a46d73fdabb21524399983855041311a4324fa93f9acd2b8fcfa51158ce8b031","correctionReason":null,"id":"7a84a12fb5bfd31974109de52d19c361efc8f23695076ff401d43497cd638a91","publicationId":"9f05af90-ea85-4d26-8130-c7c88e9ea2fa","version":1,"sourceUrl":"https://arxiv.org/abs/2611.11995","sourceTitle":"Infrathon: Benchmarking Language Agents for Real-World, Long-Horizon AI Infra Optimization","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-09","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-10T08:39:54.499Z","createdAt":"2026-11-12T09:47:26.649Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"a46d73fdabb21524399983855041311a4324fa93f9acd2b8fcfa51158ce8b031","kind":"article","url":"https://arxiv.org/abs/2611.11995","sourceUrl":"https://arxiv.org/abs/2611.11995","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-12T09:46:39.304Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.11995","checkedAt":"2026-11-12T10:50:08.144Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v6","mediaId":"a46d73fdabb21524399983855041311a4324fa93f9acd2b8fcfa51158ce8b031","confidence":0.84,"provider":"typesafe","model":"jev-1.13.0","inputHash":"c4675c50e9ec84d79e9915fae37c113daf02800170be383d8a1841d75147ca58","decidedAt":"2026-11-12T09:46:40.193Z"},"provenance":{"textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observationId":"c2afaa2a003a8073d02a21d2d8fc64672c4f07b1678895965c9275dc7cbe5211","retrievedAt":"2026-11-12T09:46:39.211Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"61010c70-3916-49a0-8194-1d233e234ff5"},"reviewer":{"model":"gpt-6-luna","runId":"b779c33f-1da8-4782-8f0c-a5115e8b54a5"}},"audit":{"status":"passed","reviewedAt":"2026-11-12T09:47:26.649Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Hu, Zhengyu","url":null,"kind":"unknown"},{"name":"Xiao, Zheyuan","url":null,"kind":"unknown"},{"name":"Wang, Yike","url":null,"kind":"unknown"},{"name":"Ivison, Hamish","url":null,"kind":"unknown"},{"name":"Wang, Haoyu","url":null,"kind":"unknown"},{"name":"Yuan, Yige","url":null,"kind":"unknown"},{"name":"Shao, Rulin","url":null,"kind":"unknown"},{"name":"Li, Yuetai","url":null,"kind":"unknown"},{"name":"Lyu, Niqi","url":null,"kind":"unknown"},{"name":"Ding, Yicheng","url":null,"kind":"unknown"},{"name":"Yang, Shu","url":null,"kind":"unknown"},{"name":"Zeng, Zhichen","url":null,"kind":"unknown"},{"name":"Zhu, Banghua","url":null,"kind":"unknown"},{"name":"Pei, Jiaxin","url":null,"kind":"unknown"},{"name":"Ding, Kaize","url":null,"kind":"unknown"},{"name":"Dasigi, Pradeep","url":null,"kind":"unknown"},{"name":"Roth, Dan","url":null,"kind":"unknown"},{"name":"Hajishirzi, Hannaneh","url":null,"kind":"unknown"},{"name":"Smith, Noah A.","url":null,"kind":"unknown"},{"name":"Xiao, Teng","url":null,"kind":"unknown"},{"name":"Poovendran, Radha","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Hu, Zhengyu","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Hu, Zhengyu","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Xiao, Zheyuan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Xiao, Zheyuan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Wang, Yike","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Wang, Yike","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Ivison, Hamish","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Ivison, Hamish","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Wang, Haoyu","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Wang, Haoyu","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:4","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Yuan, Yige","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Yuan, Yige","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:5","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Shao, Rulin","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Shao, Rulin","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:6","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Li, Yuetai","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Li, Yuetai","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:7","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Lyu, Niqi","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Lyu, Niqi","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:8","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Ding, Yicheng","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Ding, Yicheng","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:9","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Yang, Shu","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Yang, Shu","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:10","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Zeng, Zhichen","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zeng, Zhichen","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:11","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Zhu, Banghua","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhu, Banghua","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:12","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Pei, Jiaxin","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Pei, Jiaxin","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:13","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Ding, Kaize","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Ding, Kaize","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:14","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Dasigi, Pradeep","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Dasigi, Pradeep","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:15","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Roth, Dan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Roth, Dan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:16","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Hajishirzi, Hannaneh","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Hajishirzi, Hannaneh","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:17","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Smith, Noah A.","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Smith, Noah A.","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:18","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Xiao, Teng","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Xiao, Teng","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:19","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Poovendran, Radha","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Poovendran, Radha","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:20","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}}],"participationCoverage":{"total":21,"returned":21,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.11995","snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z","basis":"publisher-page-metadata","sourceLinks":60,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/7a84a12fb5bfd31974109de52d19c361efc8f23695076ff401d43497cd638a91/source"}},"canonicalUrl":"https://escapemeridian.com/readings/infrathon-the-authors-test-language-agents-on-long-horizon-ai-7a84a12f","formats":{"html":"/readings/infrathon-the-authors-test-language-agents-on-long-horizon-ai-7a84a12f","markdown":"/readings/infrathon-the-authors-test-language-agents-on-long-horizon-ai-7a84a12f.md","json":"/readings/infrathon-the-authors-test-language-agents-on-long-horizon-ai-7a84a12f.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[{"id":"term-benchmark","name":"Benchmark","explanation":"A defined set of tests used to compare systems. Results depend on the tasks and measures selected.","referenceUrl":"https://arxiv.org/abs/2211.09110"}],"entities":[{"id":"arxiv","name":"arXiv","kind":"organization","context":"arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.","referenceUrls":["https://arxiv.org/","https://info.arxiv.org/help/api/index.html"],"path":"/entities/arxiv","json":"/entities/arxiv.json","markdown":"/entities/arxiv.md"},{"id":"claude","name":"Claude","kind":"model","context":"Claude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding.","referenceUrls":["https://claude.com/product/overview","https://www.anthropic.com/news","https://www.anthropic.com/research/team/economics"],"path":"/entities/claude","json":"/entities/claude.json","markdown":"/entities/claude.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"7a84a12fb5bfd31974109de52d19c361efc8f23695076ff401d43497cd638a91","revision":"9f05af90-ea85-4d26-8130-c7c88e9ea2fa","revisionKind":"publication-id","title":"Infrathon: The authors test language agents on long-horizon AI infrastructure optimization","url":"https://escapemeridian.com/readings/infrathon-the-authors-test-language-agents-on-long-horizon-ai-7a84a12f"},"source":{"type":"source","id":"https://arxiv.org/abs/2611.11995","revision":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","revisionKind":"sha256-source-text","title":"Infrathon: Benchmarking Language Agents for Real-World, Long-Horizon AI Infra Optimization","url":"https://arxiv.org/abs/2611.11995"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"7a84a12fb5bfd31974109de52d19c361efc8f23695076ff401d43497cd638a91","title":"Infrathon: The authors test language agents on long-horizon AI infrastructure optimization","argument":"The authors of Infrathon ask whether frontier agents can keep improving real-world AI infrastructure over extended work periods. Their benchmark covers diverse AI infrastructure optimization tasks, and they evaluate six models across 150 runs with 24-hour budgets. They report that model rankings vary with the evaluation horizon, and that 135 runs improved further in the second half of their budgets. In their results, shorter evaluations can miss both later gains and shifts in which model performs best. The authors also report a case where Claude Opus 5.5 raised the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%. The comparison suggests agents may find infrastructure solutions that outperform human-designed baselines on a measured task; it does not establish that result across other tasks or deployments. The authors identify task violations, unreliable testing and premature completion as barriers to sustained improvement. The abstract presents Infrathon as a resource for evaluating agents over longer periods and discovering infrastructure solutions.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The supplied evidence is the paper's abstract, which does not detail the benchmark tasks, evaluation methods, or how the reported barriers affected individual runs. The reported codec comparison is for one stated task; the abstract does not establish performance across other tasks or real-world deployments.","speaker":{"name":"Hu, Zhengyu, Xiao, Zheyuan, Wang, Yike, Ivison, Hamish, Wang, Haoyu, Yuan, Yige, Shao, Rulin, Li, Yuetai, Lyu, Niqi, Ding, Yicheng, Yang, Shu, Zeng, Zhichen, Zhu, Banghua, Pei, Jiaxin, Ding, Kaize, Dasigi, Pradeep, Roth, Dan, Hajishirzi, Hannaneh, Smith, Noah A., Xiao, Teng, Poovendran, Radha","roleAtSource":"Article author"},"source":{"title":"Infrathon: Benchmarking Language Agents for Real-World, Long-Horizon AI Infra Optimization","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.11995","publishedAt":"2026-11-09","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors introduce Infrathon to study whether frontier agents can continually improve real-world AI infrastructure over long horizons.","excerpt":"In this work, we study whether frontier agent can continually improve real-world AI infrastructure over long horizons. We introduce Infrathon","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"The authors evaluate six models in 150 runs, each with a 24-hour budget, totaling over 10,000 GPU-hours.","excerpt":"We evaluate six models across 150 runs with 24-hour budgets, totaling over 10,000 GPU-hours.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"Model rankings vary with evaluation horizon, and 135 of 150 runs improved further during the second half of their budgets.","excerpt":"model rankings vary with the evaluation horizon, while 135 of the 150 runs achieve further improvements during the second half of their budgets.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"The authors say shorter evaluations can miss both optimization gains and changes in relative model performance.","excerpt":"These findings indicate that shorter evaluations can overlook both optimization gains and changes in relative model performance.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"The abstract reports Claude Opus 5.5 increased mean KV-cache compression ratio from 6.82% with a human-designed codec to 19.86%.","excerpt":"Claude Opus 5.5 increases the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"},{"text":"The authors identify task violations, unreliable testing and premature completion as barriers to sustained improvement.","excerpt":"Analysis of agent trajectories identifies task violations, unreliable testing, and premature completion as barriers to sustained improvement.","url":"https://arxiv.org/abs/2611.11995","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.11995","checkedAt":"2026-11-12T10:50:08.144Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How much does each of task violations, unreliable testing and premature completion limit sustained improvement in the benchmark runs?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"Infrathon: The authors test language agents on long-horizon AI infrastructure optimization","summary":"The authors of Infrathon ask whether frontier agents can keep improving real-world AI infrastructure over extended work periods. Their benchmark covers diverse AI infrastructure optimization tasks, and they evaluate six models across 150 runs with 24-hour budgets. They report that model rankings vary with the evaluation horizon, and that 135 runs improved further in the second half of their budgets. In their results, shorter evaluations can miss both later gains and shifts in which model performs best. The authors also report a case where Claude Opus 5.5 raised the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%. The comparison suggests agents may find infrastructure solutions that outperform human-designed baselines on a measured task; it does not establish that result across other tasks or deployments. The authors identify task violations, unreliable testing and premature completion as barriers to sustained improvement. The abstract presents Infrathon as a resource for evaluating agents over longer periods and discovering infrastructure solutions.","question":"How much does each of task violations, unreliable testing and premature completion limit sustained improvement in the benchmark runs?","claims":[{"text":"The authors introduce Infrathon to study whether frontier agents can continually improve real-world AI infrastructure over long horizons.","excerpt":"In this work, we study whether frontier agent can continually improve real-world AI infrastructure over long horizons. We introduce Infrathon","support":"source-stated"},{"text":"The authors evaluate six models in 150 runs, each with a 24-hour budget, totaling over 10,000 GPU-hours.","excerpt":"We evaluate six models across 150 runs with 24-hour budgets, totaling over 10,000 GPU-hours.","support":"source-stated"},{"text":"Model rankings vary with evaluation horizon, and 135 of 150 runs improved further during the second half of their budgets.","excerpt":"model rankings vary with the evaluation horizon, while 135 of the 150 runs achieve further improvements during the second half of their budgets.","support":"source-stated"},{"text":"The authors say shorter evaluations can miss both optimization gains and changes in relative model performance.","excerpt":"These findings indicate that shorter evaluations can overlook both optimization gains and changes in relative model performance.","support":"source-stated"},{"text":"The abstract reports Claude Opus 5.5 increased mean KV-cache compression ratio from 6.82% with a human-designed codec to 19.86%.","excerpt":"Claude Opus 5.5 increases the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%.","support":"source-stated"},{"text":"The authors identify task violations, unreliable testing and premature completion as barriers to sustained improvement.","excerpt":"Analysis of agent trajectories identifies task violations, unreliable testing, and premature completion as barriers to sustained improvement.","support":"source-stated"}],"limitations":["The supplied evidence is the paper's abstract, which does not detail the benchmark tasks, evaluation methods, or how the reported barriers affected individual runs.","The reported codec comparison is for one stated task; the abstract does not establish performance across other tasks or real-world deployments."],"mediaId":"a46d73fdabb21524399983855041311a4324fa93f9acd2b8fcfa51158ce8b031","correctionReason":null,"id":"7a84a12fb5bfd31974109de52d19c361efc8f23695076ff401d43497cd638a91","publicationId":"9f05af90-ea85-4d26-8130-c7c88e9ea2fa","version":1,"sourceUrl":"https://arxiv.org/abs/2611.11995","sourceTitle":"Infrathon: Benchmarking Language Agents for Real-World, Long-Horizon AI Infra Optimization","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-09","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-10T08:39:54.499Z","createdAt":"2026-11-12T09:47:26.649Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"a46d73fdabb21524399983855041311a4324fa93f9acd2b8fcfa51158ce8b031","kind":"article","url":"https://arxiv.org/abs/2611.11995","sourceUrl":"https://arxiv.org/abs/2611.11995","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-12T09:46:39.304Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.11995","checkedAt":"2026-11-12T10:50:08.144Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v6","mediaId":"a46d73fdabb21524399983855041311a4324fa93f9acd2b8fcfa51158ce8b031","confidence":0.84,"provider":"typesafe","model":"jev-1.13.0","inputHash":"c4675c50e9ec84d79e9915fae37c113daf02800170be383d8a1841d75147ca58","decidedAt":"2026-11-12T09:46:40.193Z"},"provenance":{"textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observationId":"c2afaa2a003a8073d02a21d2d8fc64672c4f07b1678895965c9275dc7cbe5211","retrievedAt":"2026-11-12T09:46:39.211Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"61010c70-3916-49a0-8194-1d233e234ff5"},"reviewer":{"model":"gpt-6-luna","runId":"b779c33f-1da8-4782-8f0c-a5115e8b54a5"}},"audit":{"status":"passed","reviewedAt":"2026-11-12T09:47:26.649Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Hu, Zhengyu","url":null,"kind":"unknown"},{"name":"Xiao, Zheyuan","url":null,"kind":"unknown"},{"name":"Wang, Yike","url":null,"kind":"unknown"},{"name":"Ivison, Hamish","url":null,"kind":"unknown"},{"name":"Wang, Haoyu","url":null,"kind":"unknown"},{"name":"Yuan, Yige","url":null,"kind":"unknown"},{"name":"Shao, Rulin","url":null,"kind":"unknown"},{"name":"Li, Yuetai","url":null,"kind":"unknown"},{"name":"Lyu, Niqi","url":null,"kind":"unknown"},{"name":"Ding, Yicheng","url":null,"kind":"unknown"},{"name":"Yang, Shu","url":null,"kind":"unknown"},{"name":"Zeng, Zhichen","url":null,"kind":"unknown"},{"name":"Zhu, Banghua","url":null,"kind":"unknown"},{"name":"Pei, Jiaxin","url":null,"kind":"unknown"},{"name":"Ding, Kaize","url":null,"kind":"unknown"},{"name":"Dasigi, Pradeep","url":null,"kind":"unknown"},{"name":"Roth, Dan","url":null,"kind":"unknown"},{"name":"Hajishirzi, Hannaneh","url":null,"kind":"unknown"},{"name":"Smith, Noah A.","url":null,"kind":"unknown"},{"name":"Xiao, Teng","url":null,"kind":"unknown"},{"name":"Poovendran, Radha","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Hu, Zhengyu","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Hu, Zhengyu","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Xiao, Zheyuan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Xiao, Zheyuan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Wang, Yike","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Wang, Yike","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Ivison, Hamish","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Ivison, Hamish","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Wang, Haoyu","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Wang, Haoyu","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:4","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Yuan, Yige","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Yuan, Yige","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:5","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Shao, Rulin","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Shao, Rulin","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:6","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Li, Yuetai","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Li, Yuetai","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:7","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Lyu, Niqi","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Lyu, Niqi","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:8","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Ding, Yicheng","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Ding, Yicheng","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:9","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Yang, Shu","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Yang, Shu","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:10","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Zeng, Zhichen","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zeng, Zhichen","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:11","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Zhu, Banghua","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhu, Banghua","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:12","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Pei, Jiaxin","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Pei, Jiaxin","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:13","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Ding, Kaize","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Ding, Kaize","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:14","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Dasigi, Pradeep","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Dasigi, Pradeep","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:15","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Roth, Dan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Roth, Dan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:16","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Hajishirzi, Hannaneh","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Hajishirzi, Hannaneh","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:17","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Smith, Noah A.","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Smith, Noah A.","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:18","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Xiao, Teng","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Xiao, Teng","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:19","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}},{"actor":{"name":"Poovendran, Radha","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Poovendran, Radha","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.11995","locator":"head/meta[name=author|citation_author]"},"assertionId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b:participation:20","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T10:50:08.144Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","sourceUrl":"https://arxiv.org/abs/2611.11995","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z"}}],"participationCoverage":{"total":21,"returned":21,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.11995","snapshotId":"a0cd8569797b4c1a69054fb64004c71e6a87da40dd5248c7537d6df04bb24b6b","contentSha256":"ca4b4897672263b9fb8e55e38e7d18d6039fc8b615e5b43aba9d83425914a408","textHash":"61a0f9c9c14bac2746e47a97c7fba725c27eb5f52894ad47c2eb33228fd0329c","observedAt":"2026-11-12T10:50:08.144Z","basis":"publisher-page-metadata","sourceLinks":60,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/7a84a12fb5bfd31974109de52d19c361efc8f23695076ff401d43497cd638a91/source"}}}],"audits":[{"publicationId":"9f05af90-ea85-4d26-8130-c7c88e9ea2fa","verdict":"passed","findings":[{"dimension":"attribution","status":"pass","explanation":"The draft consistently attributes the benchmark, results and interpretations to the paper's authors or abstract. The arXiv authorship and the Claude Opus 5.5 result are presented as reported claims, not as established facts.","excerpt":"We introduce Infrathon, a benchmark of diverse, long-horizon AI infrastructure optimization tasks across different levels."},{"dimension":"summary","status":"pass","explanation":"The title, summary and question match the abstract. The reading leads with what the authors test, then explains why it matters: shorter evaluations can miss later gains and shifts in model rankings. The question about the relative weight of the three named barriers is a genuine open issue. KV-cache compression ratio is left unexplained, but it is a supporting example rather than the central development. A term-library tooltip for KV-cache would help readers.","excerpt":"These findings indicate that shorter evaluations can overlook both optimization gains and changes in relative model performance."},{"dimension":"date","status":"pass","explanation":"The original publication date of 2026-11-09 matches the arXiv v1 submission date. It is not the later feed date.","excerpt":"[Submitted on 9 Nov 2026]"},{"dimension":"support","status":"pass","explanation":"All six claims are verbatim-supported. The hedge on the codec result stays within the source and does not overstate it: the result is limited to one measured task and does not establish deployment performance. The draft invents no benefits, costs or stakeholders.","excerpt":"Claude Opus 5.5 increases the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%."},{"dimension":"media","status":"pass","explanation":"The selected media is the verified, direct arXiv abstract page for this paper, which is the original source.","excerpt":null}],"createdAt":"2026-11-12T09:47:26.649Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}