{"id":"8b6552da8ccbea89a0da2afd707997a5f94df232f17494996cda1642ab9f63b5","title":"arXiv cs.AI: GREP: Grounding Generated Room Evaluation in Principles","argument":"The authors of GREP argue that evaluating furnished rooms generated by selecting and arranging assets requires more than checking a few perspectives or preset questions. They propose three principles: basic physical validity checks whether a scene obeys physical constraints; visual coherence concerns compatibility in object scale, style, and composition; and human-use logic asks whether arrangements fit everyday human habits. An evaluation agent uses these principles to identify scene-specific issues, gather evidence, and produce scores and reports. The authors report 87% agreement with human majority preferences, and say GREP reports improved scene quality through refinement, unlike score-only and no-feedback baselines, which showed slight declines. They also report that as planning ability improves, recent models such as GPT-5.6 Sol can challenge the guidance and constraints of pipelines and harnesses, but still lack the ability to resolve resulting conflicts. The authors argue that stronger models will need to adapt more flexibly to external guidance and constraints. The abstract presents these results but does not give the experimental details needed to assess their scope.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The supplied evidence is the paper's abstract; it does not provide experimental details needed to assess the scope of the reported results.","speaker":{"name":"Cao, Mohan, Mao, Yongsen, Xiang, Chucheng, Liu, Zhongyuan, Zhong, Zhihang, Wang, Haiping, Liu, Yuan","roleAtSource":"Article author"},"source":{"title":"GREP: Grounding Generated Room Evaluation in Principles","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.05802","publishedAt":"2026-11-04","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors argue that room evaluation should use three principles: physical validity, visual coherence, and human-use logic.","excerpt":"a three-tier evaluation philosophy: basic physical validity checks whether a scene obeys physical constraints; visual coherence concerns compatibility in object scale, style, and composition; and, at the highest level, human-use logic","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"},{"text":"GREP's evaluation agent identifies scene-specific issues, gathers evidence, and produces scores and reports.","excerpt":"an evaluation agent autonomously and adaptively identifies issues from scene context, gathers evidence, and produces scores and reports","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"},{"text":"The authors report that GREP achieves 87% agreement with human majority preferences and outperforms baseline evaluators.","excerpt":"GREP achieves 87% agreement with human majority preferences, substantially outperforming baseline evaluators.","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"},{"text":"The authors report that refinement using GREP reports improved scene quality, while score-only and no-feedback baselines showed slight declines.","excerpt":"GREP reports improve scene quality through refinement, whereas score-only and no-feedback baselines show slight declines.","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"},{"text":"The authors say recent models can challenge guidance and constraints as planning ability improves, but still lack the ability to resolve resulting conflicts.","excerpt":"as planning ability improves, recent models such as GPT-5.6 Sol can move from relying on pipelines and harnesses to challenging their guidance and constraints, but still lack the ability to resolve the resulting conflicts.","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.05802","checkedAt":"2026-11-09T08:50:08.403Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"Across what kinds of generated room scenes does GREP achieve 87% agreement with human majority preferences?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"arXiv cs.AI: GREP: Grounding Generated Room Evaluation in Principles","summary":"The authors of GREP argue that evaluating furnished rooms generated by selecting and arranging assets requires more than checking a few perspectives or preset questions. They propose three principles: basic physical validity checks whether a scene obeys physical constraints; visual coherence concerns compatibility in object scale, style, and composition; and human-use logic asks whether arrangements fit everyday human habits. An evaluation agent uses these principles to identify scene-specific issues, gather evidence, and produce scores and reports. The authors report 87% agreement with human majority preferences, and say GREP reports improved scene quality through refinement, unlike score-only and no-feedback baselines, which showed slight declines. They also report that as planning ability improves, recent models such as GPT-5.6 Sol can challenge the guidance and constraints of pipelines and harnesses, but still lack the ability to resolve resulting conflicts. The authors argue that stronger models will need to adapt more flexibly to external guidance and constraints. The abstract presents these results but does not give the experimental details needed to assess their scope.","question":"Across what kinds of generated room scenes does GREP achieve 87% agreement with human majority preferences?","claims":[{"text":"The authors argue that room evaluation should use three principles: physical validity, visual coherence, and human-use logic.","excerpt":"a three-tier evaluation philosophy: basic physical validity checks whether a scene obeys physical constraints; visual coherence concerns compatibility in object scale, style, and composition; and, at the highest level, human-use logic","support":"source-stated"},{"text":"GREP's evaluation agent identifies scene-specific issues, gathers evidence, and produces scores and reports.","excerpt":"an evaluation agent autonomously and adaptively identifies issues from scene context, gathers evidence, and produces scores and reports","support":"source-stated"},{"text":"The authors report that GREP achieves 87% agreement with human majority preferences and outperforms baseline evaluators.","excerpt":"GREP achieves 87% agreement with human majority preferences, substantially outperforming baseline evaluators.","support":"source-stated"},{"text":"The authors report that refinement using GREP reports improved scene quality, while score-only and no-feedback baselines showed slight declines.","excerpt":"GREP reports improve scene quality through refinement, whereas score-only and no-feedback baselines show slight declines.","support":"source-stated"},{"text":"The authors say recent models can challenge guidance and constraints as planning ability improves, but still lack the ability to resolve resulting conflicts.","excerpt":"as planning ability improves, recent models such as GPT-5.6 Sol can move from relying on pipelines and harnesses to challenging their guidance and constraints, but still lack the ability to resolve the resulting conflicts.","support":"source-stated"}],"limitations":["The supplied evidence is the paper's abstract; it does not provide experimental details needed to assess the scope of the reported results."],"mediaId":"9b09ec612b8970da67f895c7dad0dfae243034e84cf50a12ee2dc43107116b46","correctionReason":null,"id":"8b6552da8ccbea89a0da2afd707997a5f94df232f17494996cda1642ab9f63b5","publicationId":"42ef6fc4-c509-43d5-8f27-94ddfaa7fd45","version":1,"sourceUrl":"https://arxiv.org/abs/2611.05802","sourceTitle":"GREP: Grounding Generated Room Evaluation in Principles","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-04","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-05T05:06:34.885Z","createdAt":"2026-11-09T08:18:10.365Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"9b09ec612b8970da67f895c7dad0dfae243034e84cf50a12ee2dc43107116b46","kind":"article","url":"https://arxiv.org/abs/2611.05802","sourceUrl":"https://arxiv.org/abs/2611.05802","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-09T08:17:21.729Z","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/2611.05802","checkedAt":"2026-11-09T08:50:08.403Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"9b09ec612b8970da67f895c7dad0dfae243034e84cf50a12ee2dc43107116b46","confidence":0.8300000000000001,"provider":"typesafe","model":"jev-1.13.0","inputHash":"955624196a05ab5c49f5331e871090e460682cbd563a43ea4b40207ec77d0da2","decidedAt":"2026-11-09T08:17:22.723Z"},"provenance":{"textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observationId":"abecf28e56104ca4d520b873ceb4cbdd411ddbf0d09501922e483caced368810","retrievedAt":"2026-11-09T08:17:21.566Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"caa365ed-8831-43bc-8244-90ec6015d8fe"},"reviewer":{"model":"gpt-6-luna","runId":"2c40d005-c4a6-4167-82bd-f0368fcb4013"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-09T08:18:10.365Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Cao, Mohan","url":null,"kind":"unknown"},{"name":"Mao, Yongsen","url":null,"kind":"unknown"},{"name":"Xiang, Chucheng","url":null,"kind":"unknown"},{"name":"Liu, Zhongyuan","url":null,"kind":"unknown"},{"name":"Zhong, Zhihang","url":null,"kind":"unknown"},{"name":"Wang, Haiping","url":null,"kind":"unknown"},{"name":"Liu, Yuan","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Cao, Mohan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Cao, Mohan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Mao, Yongsen","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Mao, Yongsen","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Xiang, Chucheng","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Xiang, Chucheng","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Liu, Zhongyuan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Liu, Zhongyuan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Zhong, Zhihang","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhong, Zhihang","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:4","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Wang, Haiping","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Wang, Haiping","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:5","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Liu, Yuan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Liu, Yuan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:6","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}}],"participationCoverage":{"total":7,"returned":7,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.05802","snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z","basis":"publisher-page-metadata","sourceLinks":48,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/8b6552da8ccbea89a0da2afd707997a5f94df232f17494996cda1642ab9f63b5/source"}},"canonicalUrl":"https://escapemeridian.com/readings/arxiv-cs-ai-grep-grounding-generated-room-evaluation-in-8b6552da","formats":{"html":"/readings/arxiv-cs-ai-grep-grounding-generated-room-evaluation-in-8b6552da","markdown":"/readings/arxiv-cs-ai-grep-grounding-generated-room-evaluation-in-8b6552da.md","json":"/readings/arxiv-cs-ai-grep-grounding-generated-room-evaluation-in-8b6552da.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":[],"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":"8b6552da8ccbea89a0da2afd707997a5f94df232f17494996cda1642ab9f63b5","revision":"42ef6fc4-c509-43d5-8f27-94ddfaa7fd45","revisionKind":"publication-id","title":"arXiv cs.AI: GREP: Grounding Generated Room Evaluation in Principles","url":"https://escapemeridian.com/readings/arxiv-cs-ai-grep-grounding-generated-room-evaluation-in-8b6552da"},"source":{"type":"source","id":"https://arxiv.org/abs/2611.05802","revision":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","revisionKind":"sha256-source-text","title":"GREP: Grounding Generated Room Evaluation in Principles","url":"https://arxiv.org/abs/2611.05802"},"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":"8b6552da8ccbea89a0da2afd707997a5f94df232f17494996cda1642ab9f63b5","title":"arXiv cs.AI: GREP: Grounding Generated Room Evaluation in Principles","argument":"The authors of GREP argue that evaluating furnished rooms generated by selecting and arranging assets requires more than checking a few perspectives or preset questions. They propose three principles: basic physical validity checks whether a scene obeys physical constraints; visual coherence concerns compatibility in object scale, style, and composition; and human-use logic asks whether arrangements fit everyday human habits. An evaluation agent uses these principles to identify scene-specific issues, gather evidence, and produce scores and reports. The authors report 87% agreement with human majority preferences, and say GREP reports improved scene quality through refinement, unlike score-only and no-feedback baselines, which showed slight declines. They also report that as planning ability improves, recent models such as GPT-5.6 Sol can challenge the guidance and constraints of pipelines and harnesses, but still lack the ability to resolve resulting conflicts. The authors argue that stronger models will need to adapt more flexibly to external guidance and constraints. The abstract presents these results but does not give the experimental details needed to assess their scope.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The supplied evidence is the paper's abstract; it does not provide experimental details needed to assess the scope of the reported results.","speaker":{"name":"Cao, Mohan, Mao, Yongsen, Xiang, Chucheng, Liu, Zhongyuan, Zhong, Zhihang, Wang, Haiping, Liu, Yuan","roleAtSource":"Article author"},"source":{"title":"GREP: Grounding Generated Room Evaluation in Principles","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.05802","publishedAt":"2026-11-04","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors argue that room evaluation should use three principles: physical validity, visual coherence, and human-use logic.","excerpt":"a three-tier evaluation philosophy: basic physical validity checks whether a scene obeys physical constraints; visual coherence concerns compatibility in object scale, style, and composition; and, at the highest level, human-use logic","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"},{"text":"GREP's evaluation agent identifies scene-specific issues, gathers evidence, and produces scores and reports.","excerpt":"an evaluation agent autonomously and adaptively identifies issues from scene context, gathers evidence, and produces scores and reports","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"},{"text":"The authors report that GREP achieves 87% agreement with human majority preferences and outperforms baseline evaluators.","excerpt":"GREP achieves 87% agreement with human majority preferences, substantially outperforming baseline evaluators.","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"},{"text":"The authors report that refinement using GREP reports improved scene quality, while score-only and no-feedback baselines showed slight declines.","excerpt":"GREP reports improve scene quality through refinement, whereas score-only and no-feedback baselines show slight declines.","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"},{"text":"The authors say recent models can challenge guidance and constraints as planning ability improves, but still lack the ability to resolve resulting conflicts.","excerpt":"as planning ability improves, recent models such as GPT-5.6 Sol can move from relying on pipelines and harnesses to challenging their guidance and constraints, but still lack the ability to resolve the resulting conflicts.","url":"https://arxiv.org/abs/2611.05802","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.05802","checkedAt":"2026-11-09T08:50:08.403Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"Across what kinds of generated room scenes does GREP achieve 87% agreement with human majority preferences?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"arXiv cs.AI: GREP: Grounding Generated Room Evaluation in Principles","summary":"The authors of GREP argue that evaluating furnished rooms generated by selecting and arranging assets requires more than checking a few perspectives or preset questions. They propose three principles: basic physical validity checks whether a scene obeys physical constraints; visual coherence concerns compatibility in object scale, style, and composition; and human-use logic asks whether arrangements fit everyday human habits. An evaluation agent uses these principles to identify scene-specific issues, gather evidence, and produce scores and reports. The authors report 87% agreement with human majority preferences, and say GREP reports improved scene quality through refinement, unlike score-only and no-feedback baselines, which showed slight declines. They also report that as planning ability improves, recent models such as GPT-5.6 Sol can challenge the guidance and constraints of pipelines and harnesses, but still lack the ability to resolve resulting conflicts. The authors argue that stronger models will need to adapt more flexibly to external guidance and constraints. The abstract presents these results but does not give the experimental details needed to assess their scope.","question":"Across what kinds of generated room scenes does GREP achieve 87% agreement with human majority preferences?","claims":[{"text":"The authors argue that room evaluation should use three principles: physical validity, visual coherence, and human-use logic.","excerpt":"a three-tier evaluation philosophy: basic physical validity checks whether a scene obeys physical constraints; visual coherence concerns compatibility in object scale, style, and composition; and, at the highest level, human-use logic","support":"source-stated"},{"text":"GREP's evaluation agent identifies scene-specific issues, gathers evidence, and produces scores and reports.","excerpt":"an evaluation agent autonomously and adaptively identifies issues from scene context, gathers evidence, and produces scores and reports","support":"source-stated"},{"text":"The authors report that GREP achieves 87% agreement with human majority preferences and outperforms baseline evaluators.","excerpt":"GREP achieves 87% agreement with human majority preferences, substantially outperforming baseline evaluators.","support":"source-stated"},{"text":"The authors report that refinement using GREP reports improved scene quality, while score-only and no-feedback baselines showed slight declines.","excerpt":"GREP reports improve scene quality through refinement, whereas score-only and no-feedback baselines show slight declines.","support":"source-stated"},{"text":"The authors say recent models can challenge guidance and constraints as planning ability improves, but still lack the ability to resolve resulting conflicts.","excerpt":"as planning ability improves, recent models such as GPT-5.6 Sol can move from relying on pipelines and harnesses to challenging their guidance and constraints, but still lack the ability to resolve the resulting conflicts.","support":"source-stated"}],"limitations":["The supplied evidence is the paper's abstract; it does not provide experimental details needed to assess the scope of the reported results."],"mediaId":"9b09ec612b8970da67f895c7dad0dfae243034e84cf50a12ee2dc43107116b46","correctionReason":null,"id":"8b6552da8ccbea89a0da2afd707997a5f94df232f17494996cda1642ab9f63b5","publicationId":"42ef6fc4-c509-43d5-8f27-94ddfaa7fd45","version":1,"sourceUrl":"https://arxiv.org/abs/2611.05802","sourceTitle":"GREP: Grounding Generated Room Evaluation in Principles","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-04","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-05T05:06:34.885Z","createdAt":"2026-11-09T08:18:10.365Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"9b09ec612b8970da67f895c7dad0dfae243034e84cf50a12ee2dc43107116b46","kind":"article","url":"https://arxiv.org/abs/2611.05802","sourceUrl":"https://arxiv.org/abs/2611.05802","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-09T08:17:21.729Z","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/2611.05802","checkedAt":"2026-11-09T08:50:08.403Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"9b09ec612b8970da67f895c7dad0dfae243034e84cf50a12ee2dc43107116b46","confidence":0.8300000000000001,"provider":"typesafe","model":"jev-1.13.0","inputHash":"955624196a05ab5c49f5331e871090e460682cbd563a43ea4b40207ec77d0da2","decidedAt":"2026-11-09T08:17:22.723Z"},"provenance":{"textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observationId":"abecf28e56104ca4d520b873ceb4cbdd411ddbf0d09501922e483caced368810","retrievedAt":"2026-11-09T08:17:21.566Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"caa365ed-8831-43bc-8244-90ec6015d8fe"},"reviewer":{"model":"gpt-6-luna","runId":"2c40d005-c4a6-4167-82bd-f0368fcb4013"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-09T08:18:10.365Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Cao, Mohan","url":null,"kind":"unknown"},{"name":"Mao, Yongsen","url":null,"kind":"unknown"},{"name":"Xiang, Chucheng","url":null,"kind":"unknown"},{"name":"Liu, Zhongyuan","url":null,"kind":"unknown"},{"name":"Zhong, Zhihang","url":null,"kind":"unknown"},{"name":"Wang, Haiping","url":null,"kind":"unknown"},{"name":"Liu, Yuan","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Cao, Mohan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Cao, Mohan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Mao, Yongsen","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Mao, Yongsen","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Xiang, Chucheng","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Xiang, Chucheng","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Liu, Zhongyuan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Liu, Zhongyuan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Zhong, Zhihang","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhong, Zhihang","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:4","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Wang, Haiping","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Wang, Haiping","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:5","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}},{"actor":{"name":"Liu, Yuan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Liu, Yuan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.05802","locator":"head/meta[name=author|citation_author]"},"assertionId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757:participation:6","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T08:50:08.403Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","sourceUrl":"https://arxiv.org/abs/2611.05802","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z"}}],"participationCoverage":{"total":7,"returned":7,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.05802","snapshotId":"0487c00e5303f04a7a30c37c94f6a1175d6d2d306367533f896e7ee4a8420757","contentSha256":"27c95caa04d95fcb5bbca8ee5114339517b76f4362e4cf4c7d8872e50db7101d","textHash":"e7b88a751e4b726f74d4fd017ef9d42ba16899d9d729f2f986171d41b9cb581e","observedAt":"2026-11-09T08:50:08.403Z","basis":"publisher-page-metadata","sourceLinks":48,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/8b6552da8ccbea89a0da2afd707997a5f94df232f17494996cda1642ab9f63b5/source"}}}],"audits":[{"publicationId":"42ef6fc4-c509-43d5-8f27-94ddfaa7fd45","verdict":"uncertain","findings":[{"dimension":"attribution","status":"pass","explanation":"The paper and its claims are attributed to the authors. The arXiv cs.AI label matches the feed, although the paper's primary category is cs.CV with a cs.AI cross-listing. Results are consistently framed as reported by the authors.","excerpt":"Authors:Mohan Cao, Yongsen Mao, Chucheng Xiang, Zhongyuan Liu, Zhihang Zhong, Haiping Wang, Yuan Liu"},{"dimension":"summary","status":"uncertain","explanation":"The summary is faithful, but a general reader may not follow it. It never says the rooms are generated by AI (LLM-centric) systems. It leaves central terms unexplained: assets (3D object models), pipelines and harnesses (the scaffolding that guides a model), and baselines. Smallest repair: state that LLM-based systems generate virtual furnished rooms, and briefly gloss or tooltip these terms. The draft also turns the source's GREP suggests into they report, which slightly hardens that finding.","excerpt":"Recent LLM-centric advances can generate attractive furnished rooms by selecting and arranging assets."},{"dimension":"date","status":"pass","explanation":"The originalPublished value of 2026-11-04 matches the v1 submission date. It is not the later feed date.","excerpt":"[Submitted on 4 Nov 2026]"},{"dimension":"support","status":"pass","explanation":"Each claim has a matching abstract excerpt and remains attributed to the authors. The limitation correctly notes that the experimental details are missing. The question about scene scope follows from that gap without leading the reader. No benefits or costs are invented beyond the source.","excerpt":"GREP achieves 87% agreement with human majority preferences, substantially outperforming baseline evaluators."},{"dimension":"media","status":"pass","explanation":"The selected media is the verified direct link to the arXiv abstract page, which is the original source.","excerpt":null}],"createdAt":"2026-11-09T08:18:10.365Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}