{"id":"b1e6fc842634828a7cf0b0fae202102b2b0916199f514b918477854ca7cde4d0","title":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems — arXiv","argument":"Yanjie Gao, Jiayi Bi, Haoxiang Lin and Ling Feng report an empirical study of issues in GPU kernel libraries used by LLM systems. They manually analyzed 300 GitHub issue reports, discussions and available fixes across seven representative libraries. The authors find that 60.33% of the sampled issues originate in the Compiler and Build System or Runtime Environment, arguing that validation needs to cover deployment and execution as well as kernel code. They also describe how kernel specialization can impose implicit constraints on tensor shapes, metadata layouts, memory aliasing and hardware resources. Violations, they say, can cause incorrect outputs or launch failures even for supported operations. The authors argue that current testing and debugging practices often miss low-precision numerical problems and hardware-specific execution failures. Their findings make the reliability of GPU kernel libraries a practical concern for developers: problems at this layer can corrupt outputs, interrupt execution, block compilation, complicate deployment or waste GPU resources. The study’s evidence comes from selected GitHub reports in seven libraries, so its findings describe that sample rather than establish how common these issues are across all GPU kernel libraries. That final point is an Escape Meridian interpretation of the study’s sample and scope.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The study manually analyzes issue reports from seven representative libraries; the abstract does not establish how representative this sample is of issues across all GPU kernel libraries.","speaker":{"name":"Gao, Yanjie, Bi, Jiayi, Lin, Haoxiang, Feng, Ling","roleAtSource":"Article author"},"source":{"title":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.09280","publishedAt":"2026-11-06","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors manually analyzed 300 GitHub issue reports, discussions and available fixes from seven representative libraries.","excerpt":"We manually analyzed 300 high-quality GitHub issue reports, their discussions, and available fixes from seven representative libraries","url":"https://arxiv.org/abs/2611.09280","relationship":"Source-stated"},{"text":"In the sampled issues, 60.33% originate in the Compiler and Build System or Runtime Environment dimensions.","excerpt":"60.33% of the sampled issues originate in the Compiler and Build System or Runtime Environment dimensions","url":"https://arxiv.org/abs/2611.09280","relationship":"Source-stated"},{"text":"The authors say kernel specialization can impose implicit correctness constraints, and violations can cause incorrect outputs or launch failures even for supported operations.","excerpt":"Violations can produce incorrect outputs or launch failures even for supported operations","url":"https://arxiv.org/abs/2611.09280","relationship":"Source-stated"},{"text":"The authors report that current testing and debugging practices are often insufficient for identifying low-precision numerical problems and hardware-specific execution failures.","excerpt":"Current testing and debugging practices are often insufficient for identifying low-precision numerical problems and hardware-specific execution failures.","url":"https://arxiv.org/abs/2611.09280","relationship":"Source-stated"},{"text":"The sample's findings do not establish how common these issues are across all GPU kernel libraries.","excerpt":"300 high-quality GitHub issue reports, their discussions, and available fixes from seven representative libraries","url":"https://arxiv.org/abs/2611.09280","relationship":"Escape Meridian interpretation"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.09280","checkedAt":"2026-11-09T06:48:11.347Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"Which testing and debugging practices can identify low-precision numerical problems and hardware-specific execution failures in GPU kernel libraries?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems — arXiv","summary":"Yanjie Gao, Jiayi Bi, Haoxiang Lin and Ling Feng report an empirical study of issues in GPU kernel libraries used by LLM systems. They manually analyzed 300 GitHub issue reports, discussions and available fixes across seven representative libraries. The authors find that 60.33% of the sampled issues originate in the Compiler and Build System or Runtime Environment, arguing that validation needs to cover deployment and execution as well as kernel code. They also describe how kernel specialization can impose implicit constraints on tensor shapes, metadata layouts, memory aliasing and hardware resources. Violations, they say, can cause incorrect outputs or launch failures even for supported operations. The authors argue that current testing and debugging practices often miss low-precision numerical problems and hardware-specific execution failures. Their findings make the reliability of GPU kernel libraries a practical concern for developers: problems at this layer can corrupt outputs, interrupt execution, block compilation, complicate deployment or waste GPU resources. The study’s evidence comes from selected GitHub reports in seven libraries, so its findings describe that sample rather than establish how common these issues are across all GPU kernel libraries. That final point is an Escape Meridian interpretation of the study’s sample and scope.","question":"Which testing and debugging practices can identify low-precision numerical problems and hardware-specific execution failures in GPU kernel libraries?","claims":[{"text":"The authors manually analyzed 300 GitHub issue reports, discussions and available fixes from seven representative libraries.","excerpt":"We manually analyzed 300 high-quality GitHub issue reports, their discussions, and available fixes from seven representative libraries","support":"source-stated"},{"text":"In the sampled issues, 60.33% originate in the Compiler and Build System or Runtime Environment dimensions.","excerpt":"60.33% of the sampled issues originate in the Compiler and Build System or Runtime Environment dimensions","support":"source-stated"},{"text":"The authors say kernel specialization can impose implicit correctness constraints, and violations can cause incorrect outputs or launch failures even for supported operations.","excerpt":"Violations can produce incorrect outputs or launch failures even for supported operations","support":"source-stated"},{"text":"The authors report that current testing and debugging practices are often insufficient for identifying low-precision numerical problems and hardware-specific execution failures.","excerpt":"Current testing and debugging practices are often insufficient for identifying low-precision numerical problems and hardware-specific execution failures.","support":"source-stated"},{"text":"The sample's findings do not establish how common these issues are across all GPU kernel libraries.","excerpt":"300 high-quality GitHub issue reports, their discussions, and available fixes from seven representative libraries","support":"interpretation"}],"limitations":["The study manually analyzes issue reports from seven representative libraries; the abstract does not establish how representative this sample is of issues across all GPU kernel libraries."],"mediaId":"843dcbee0a11da0a8ba080b1afc59a2268bc02a24e1f86dbd9d1212d66cf072a","correctionReason":null,"id":"b1e6fc842634828a7cf0b0fae202102b2b0916199f514b918477854ca7cde4d0","publicationId":"afdf0ec0-7234-4e79-887f-198763f4d9e2","version":1,"sourceUrl":"https://arxiv.org/abs/2611.09280","sourceTitle":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-06","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-09T05:07:03.050Z","createdAt":"2026-11-09T06:16:44.521Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"843dcbee0a11da0a8ba080b1afc59a2268bc02a24e1f86dbd9d1212d66cf072a","kind":"article","url":"https://arxiv.org/abs/2611.09280","sourceUrl":"https://arxiv.org/abs/2611.09280","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-09T05:49:30.776Z","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/2611.09280","checkedAt":"2026-11-09T06:48:11.347Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["useful_perspective"],"rubricVersion":"stream-admission-v5","mediaId":"843dcbee0a11da0a8ba080b1afc59a2268bc02a24e1f86dbd9d1212d66cf072a","confidence":0.9,"provider":"typesafe","model":"jev-1.13.0","inputHash":"dbe4dbb22b7efaffc477f887fbac6ae5b6496d7a2d629557f75a9107a8d39666","decidedAt":"2026-11-09T05:49:32.529Z"},"provenance":{"textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observationId":"c3da653d5d955da0bd0140303f0e8e5c2f425757b0b0135be901e51ef9eaf520","retrievedAt":"2026-11-09T05:49:30.502Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"5bc29dc7-24a8-4a09-82cd-ae1a75607ac4"},"reviewer":{"model":"gpt-6-luna","runId":"477344ce-07cf-42ff-8e6f-b23a102b29d7"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-09T06:16:44.521Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Gao, Yanjie","url":null,"kind":"unknown"},{"name":"Bi, Jiayi","url":null,"kind":"unknown"},{"name":"Lin, Haoxiang","url":null,"kind":"unknown"},{"name":"Feng, Ling","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Gao, Yanjie","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Gao, Yanjie","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.09280","locator":"head/meta[name=author|citation_author]"},"assertionId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T06:48:11.347Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","sourceUrl":"https://arxiv.org/abs/2611.09280","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z"}},{"actor":{"name":"Bi, Jiayi","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Bi, Jiayi","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.09280","locator":"head/meta[name=author|citation_author]"},"assertionId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T06:48:11.347Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","sourceUrl":"https://arxiv.org/abs/2611.09280","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z"}},{"actor":{"name":"Lin, Haoxiang","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Lin, Haoxiang","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.09280","locator":"head/meta[name=author|citation_author]"},"assertionId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T06:48:11.347Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","sourceUrl":"https://arxiv.org/abs/2611.09280","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z"}},{"actor":{"name":"Feng, Ling","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Feng, Ling","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.09280","locator":"head/meta[name=author|citation_author]"},"assertionId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T06:48:11.347Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","sourceUrl":"https://arxiv.org/abs/2611.09280","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z"}}],"participationCoverage":{"total":4,"returned":4,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.09280","snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z","basis":"publisher-page-metadata","sourceLinks":42,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/b1e6fc842634828a7cf0b0fae202102b2b0916199f514b918477854ca7cde4d0/source"}},"canonicalUrl":"https://escapemeridian.com/readings/an-empirical-study-of-issues-in-gpu-kernel-libraries-for-llm-b1e6fc84","formats":{"html":"/readings/an-empirical-study-of-issues-in-gpu-kernel-libraries-for-llm-b1e6fc84","markdown":"/readings/an-empirical-study-of-issues-in-gpu-kernel-libraries-for-llm-b1e6fc84.md","json":"/readings/an-empirical-study-of-issues-in-gpu-kernel-libraries-for-llm-b1e6fc84.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-gpu","name":"Graphics processing unit","explanation":"A chip that performs many calculations at once; the main hardware used to train and run AI models.","referenceUrl":"https://en.wikipedia.org/wiki/Graphics_processing_unit"},{"id":"term-llm","name":"Large language model","explanation":"A language model with many learned numerical settings, used to process and generate language.","referenceUrl":"https://developers.google.com/machine-learning/glossary#large-language-model"}],"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"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"b1e6fc842634828a7cf0b0fae202102b2b0916199f514b918477854ca7cde4d0","revision":"afdf0ec0-7234-4e79-887f-198763f4d9e2","revisionKind":"publication-id","title":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems — arXiv","url":"https://escapemeridian.com/readings/an-empirical-study-of-issues-in-gpu-kernel-libraries-for-llm-b1e6fc84"},"source":{"type":"source","id":"https://arxiv.org/abs/2611.09280","revision":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","revisionKind":"sha256-source-text","title":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems","url":"https://arxiv.org/abs/2611.09280"},"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":"b1e6fc842634828a7cf0b0fae202102b2b0916199f514b918477854ca7cde4d0","title":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems — arXiv","argument":"Yanjie Gao, Jiayi Bi, Haoxiang Lin and Ling Feng report an empirical study of issues in GPU kernel libraries used by LLM systems. They manually analyzed 300 GitHub issue reports, discussions and available fixes across seven representative libraries. The authors find that 60.33% of the sampled issues originate in the Compiler and Build System or Runtime Environment, arguing that validation needs to cover deployment and execution as well as kernel code. They also describe how kernel specialization can impose implicit constraints on tensor shapes, metadata layouts, memory aliasing and hardware resources. Violations, they say, can cause incorrect outputs or launch failures even for supported operations. The authors argue that current testing and debugging practices often miss low-precision numerical problems and hardware-specific execution failures. Their findings make the reliability of GPU kernel libraries a practical concern for developers: problems at this layer can corrupt outputs, interrupt execution, block compilation, complicate deployment or waste GPU resources. The study’s evidence comes from selected GitHub reports in seven libraries, so its findings describe that sample rather than establish how common these issues are across all GPU kernel libraries. That final point is an Escape Meridian interpretation of the study’s sample and scope.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The study manually analyzes issue reports from seven representative libraries; the abstract does not establish how representative this sample is of issues across all GPU kernel libraries.","speaker":{"name":"Gao, Yanjie, Bi, Jiayi, Lin, Haoxiang, Feng, Ling","roleAtSource":"Article author"},"source":{"title":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.09280","publishedAt":"2026-11-06","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors manually analyzed 300 GitHub issue reports, discussions and available fixes from seven representative libraries.","excerpt":"We manually analyzed 300 high-quality GitHub issue reports, their discussions, and available fixes from seven representative libraries","url":"https://arxiv.org/abs/2611.09280","relationship":"Source-stated"},{"text":"In the sampled issues, 60.33% originate in the Compiler and Build System or Runtime Environment dimensions.","excerpt":"60.33% of the sampled issues originate in the Compiler and Build System or Runtime Environment dimensions","url":"https://arxiv.org/abs/2611.09280","relationship":"Source-stated"},{"text":"The authors say kernel specialization can impose implicit correctness constraints, and violations can cause incorrect outputs or launch failures even for supported operations.","excerpt":"Violations can produce incorrect outputs or launch failures even for supported operations","url":"https://arxiv.org/abs/2611.09280","relationship":"Source-stated"},{"text":"The authors report that current testing and debugging practices are often insufficient for identifying low-precision numerical problems and hardware-specific execution failures.","excerpt":"Current testing and debugging practices are often insufficient for identifying low-precision numerical problems and hardware-specific execution failures.","url":"https://arxiv.org/abs/2611.09280","relationship":"Source-stated"},{"text":"The sample's findings do not establish how common these issues are across all GPU kernel libraries.","excerpt":"300 high-quality GitHub issue reports, their discussions, and available fixes from seven representative libraries","url":"https://arxiv.org/abs/2611.09280","relationship":"Escape Meridian interpretation"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.09280","checkedAt":"2026-11-09T06:48:11.347Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"Which testing and debugging practices can identify low-precision numerical problems and hardware-specific execution failures in GPU kernel libraries?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems — arXiv","summary":"Yanjie Gao, Jiayi Bi, Haoxiang Lin and Ling Feng report an empirical study of issues in GPU kernel libraries used by LLM systems. They manually analyzed 300 GitHub issue reports, discussions and available fixes across seven representative libraries. The authors find that 60.33% of the sampled issues originate in the Compiler and Build System or Runtime Environment, arguing that validation needs to cover deployment and execution as well as kernel code. They also describe how kernel specialization can impose implicit constraints on tensor shapes, metadata layouts, memory aliasing and hardware resources. Violations, they say, can cause incorrect outputs or launch failures even for supported operations. The authors argue that current testing and debugging practices often miss low-precision numerical problems and hardware-specific execution failures. Their findings make the reliability of GPU kernel libraries a practical concern for developers: problems at this layer can corrupt outputs, interrupt execution, block compilation, complicate deployment or waste GPU resources. The study’s evidence comes from selected GitHub reports in seven libraries, so its findings describe that sample rather than establish how common these issues are across all GPU kernel libraries. That final point is an Escape Meridian interpretation of the study’s sample and scope.","question":"Which testing and debugging practices can identify low-precision numerical problems and hardware-specific execution failures in GPU kernel libraries?","claims":[{"text":"The authors manually analyzed 300 GitHub issue reports, discussions and available fixes from seven representative libraries.","excerpt":"We manually analyzed 300 high-quality GitHub issue reports, their discussions, and available fixes from seven representative libraries","support":"source-stated"},{"text":"In the sampled issues, 60.33% originate in the Compiler and Build System or Runtime Environment dimensions.","excerpt":"60.33% of the sampled issues originate in the Compiler and Build System or Runtime Environment dimensions","support":"source-stated"},{"text":"The authors say kernel specialization can impose implicit correctness constraints, and violations can cause incorrect outputs or launch failures even for supported operations.","excerpt":"Violations can produce incorrect outputs or launch failures even for supported operations","support":"source-stated"},{"text":"The authors report that current testing and debugging practices are often insufficient for identifying low-precision numerical problems and hardware-specific execution failures.","excerpt":"Current testing and debugging practices are often insufficient for identifying low-precision numerical problems and hardware-specific execution failures.","support":"source-stated"},{"text":"The sample's findings do not establish how common these issues are across all GPU kernel libraries.","excerpt":"300 high-quality GitHub issue reports, their discussions, and available fixes from seven representative libraries","support":"interpretation"}],"limitations":["The study manually analyzes issue reports from seven representative libraries; the abstract does not establish how representative this sample is of issues across all GPU kernel libraries."],"mediaId":"843dcbee0a11da0a8ba080b1afc59a2268bc02a24e1f86dbd9d1212d66cf072a","correctionReason":null,"id":"b1e6fc842634828a7cf0b0fae202102b2b0916199f514b918477854ca7cde4d0","publicationId":"afdf0ec0-7234-4e79-887f-198763f4d9e2","version":1,"sourceUrl":"https://arxiv.org/abs/2611.09280","sourceTitle":"An Empirical Study of Issues in GPU Kernel Libraries for LLM Systems","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-06","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-09T05:07:03.050Z","createdAt":"2026-11-09T06:16:44.521Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"843dcbee0a11da0a8ba080b1afc59a2268bc02a24e1f86dbd9d1212d66cf072a","kind":"article","url":"https://arxiv.org/abs/2611.09280","sourceUrl":"https://arxiv.org/abs/2611.09280","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-09T05:49:30.776Z","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/2611.09280","checkedAt":"2026-11-09T06:48:11.347Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["useful_perspective"],"rubricVersion":"stream-admission-v5","mediaId":"843dcbee0a11da0a8ba080b1afc59a2268bc02a24e1f86dbd9d1212d66cf072a","confidence":0.9,"provider":"typesafe","model":"jev-1.13.0","inputHash":"dbe4dbb22b7efaffc477f887fbac6ae5b6496d7a2d629557f75a9107a8d39666","decidedAt":"2026-11-09T05:49:32.529Z"},"provenance":{"textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observationId":"c3da653d5d955da0bd0140303f0e8e5c2f425757b0b0135be901e51ef9eaf520","retrievedAt":"2026-11-09T05:49:30.502Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"5bc29dc7-24a8-4a09-82cd-ae1a75607ac4"},"reviewer":{"model":"gpt-6-luna","runId":"477344ce-07cf-42ff-8e6f-b23a102b29d7"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-09T06:16:44.521Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Gao, Yanjie","url":null,"kind":"unknown"},{"name":"Bi, Jiayi","url":null,"kind":"unknown"},{"name":"Lin, Haoxiang","url":null,"kind":"unknown"},{"name":"Feng, Ling","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Gao, Yanjie","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Gao, Yanjie","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.09280","locator":"head/meta[name=author|citation_author]"},"assertionId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T06:48:11.347Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","sourceUrl":"https://arxiv.org/abs/2611.09280","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z"}},{"actor":{"name":"Bi, Jiayi","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Bi, Jiayi","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.09280","locator":"head/meta[name=author|citation_author]"},"assertionId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T06:48:11.347Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","sourceUrl":"https://arxiv.org/abs/2611.09280","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z"}},{"actor":{"name":"Lin, Haoxiang","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Lin, Haoxiang","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.09280","locator":"head/meta[name=author|citation_author]"},"assertionId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T06:48:11.347Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","sourceUrl":"https://arxiv.org/abs/2611.09280","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z"}},{"actor":{"name":"Feng, Ling","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Feng, Ling","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.09280","locator":"head/meta[name=author|citation_author]"},"assertionId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T06:48:11.347Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","sourceUrl":"https://arxiv.org/abs/2611.09280","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z"}}],"participationCoverage":{"total":4,"returned":4,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.09280","snapshotId":"d7abed38697d7c41240f33522e0493927b93bbe8de1ab007179684112d23f451","contentSha256":"610d9f9f4465bbc346f267713f99f9bd8fd387f3d7b5c1a37c26f7f6c03ae51e","textHash":"134cff69886bd8d80b5d5731808556cf1461731f18e4eb619480a4c8c9988c83","observedAt":"2026-11-09T06:48:11.347Z","basis":"publisher-page-metadata","sourceLinks":42,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/b1e6fc842634828a7cf0b0fae202102b2b0916199f514b918477854ca7cde4d0/source"}}}],"audits":[{"publicationId":"afdf0ec0-7234-4e79-887f-198763f4d9e2","verdict":"uncertain","findings":[{"dimension":"attribution","status":"pass","explanation":"The draft correctly names Yanjie Gao, Jiayi Bi, Haoxiang Lin and Ling Feng as authors on arXiv. It attributes the findings to them, and it labels the representativeness point as Escape Meridian interpretation.","excerpt":"Authors:Yanjie Gao, Jiayi Bi, Haoxiang Lin, Ling Feng"},{"dimension":"summary","status":"uncertain","explanation":"The central term 'GPU kernel libraries' is never explained in plain language, and neither are LLM, tensor shapes or memory aliasing. The abstract says these libraries implement core operators that underpin training and inference correctness and efficiency. Without that, a non-specialist cannot see why this layer matters. Smallest repair: add one sentence from the abstract on what these libraries do, and tooltip the terms. The draft also drops 'silently' from 'silently corrupt outputs', which weakens the source's stake.","excerpt":"These libraries underpin the correctness and efficiency of training and inference across diverse GPU platforms, implementing attention, GEMM, quantization, and low-precision operators"},{"dimension":"date","status":"pass","explanation":"The originalPublished date of 2026-11-06 matches the arXiv v1 submission date. The feed date of 2026-11-09 is later and is correctly not used.","excerpt":"[Submitted on 6 Nov 2026]"},{"dimension":"support","status":"pass","explanation":"The draft's claims match the abstract: 300 issues, seven libraries, the 60.33% figure, the constraints from kernel specialization, and the testing gaps. The consequences sentence attributes to the 'findings' what is actually the abstract's motivating framing; the harms are source-stated, but the attribution is slightly off. The scope interpretation is labeled and conditional. The question follows from finding (3).","excerpt":"Problems at this layer can silently corrupt outputs, hang GPUs, prevent compilation, introduce deployment incompatibilities, or degrade performance"},{"dimension":"media","status":"pass","explanation":"The selected media is the verified direct link to the original arXiv abstract page. It is the most direct access to the source.","excerpt":null}],"createdAt":"2026-11-09T06:16:44.521Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}