{"id":"dc794282b0c8a3665b595a22f79f507586cbafc23940cafeda2e249838675c2e","title":"arXiv cs.AI: How Should Planning Agents ReSHAPE Their Search Space?","argument":"Yilin Zhang and coauthors argue that planning agents should treat a user's request as a hypothesis, not a fixed specification. People may discover acceptable trade-offs only after seeing concrete plans, so an agent that asks about preferences without showing options can miss plans the user would accept. The authors formalize request revision as the ReSHAPE task: a revision is judged by whether its search space includes acceptable plans and a planner finds one. In ReSHAPE Planter, built from planning benchmarks with a human-validated simulated user, they analyze 5,800 agent-revised specifications in travel and shopping. They report that one correct trade-off lets the same planner find an acceptable plan up to 4.6x as often. Current agents include at most 32% of acceptable plans, compared with over 52% for that trade-off alone; in travel, showing plans raises F1 by 9.7 points over asking, with two model families. The authors recommend testing requests against users' reactions to concrete plans to recognize their trade-offs.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The abstract describes instances built from planning benchmarks with a human-validated simulated user, and reports analyses in travel and shopping; it does not establish how the results transfer to other settings or interactions with users outside the simulation. The reported comparison of showing plans with asking is specifically described for travel.","speaker":{"name":"Zhang, Yilin, Zhao, Xinran, Chen, Tong, Yuan, Ziming, Wu, Tongshuang","roleAtSource":"Article author"},"source":{"title":"How Should Planning Agents ReSHAPE Their Search Space?","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.16466","publishedAt":"2026-11-12","dateBasis":"Date basis not provided."},"evidence":[{"text":"Zhang and coauthors say users' initial requests can omit acceptable plans because people may not discover acceptable trade-offs until they see options.","excerpt":"the requests users write exclude plans they would gladly take, since they rarely discover the possible trade-offs they would accept until seeing the options.","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"},{"text":"The authors define ReSHAPE as judging whether a revised request's search space matches acceptable plans and a planner finds one there.","excerpt":"We formalize this step as the ReSHAPE task, which judges a revision by whether its search space matches the acceptable plans and a planner finds one there.","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"},{"text":"In their ReSHAPE Planter analysis of 5,800 agent-revised specifications in travel and shopping, one correct trade-off let the same planner reach an acceptable plan up to 4.6 times as often.","excerpt":"Using ReSHAPE Planter, which builds instances from planning benchmarks with a human-validated simulated user, we analyze 5,800 agent-revised specifications in travel and shopping. We find that reshaping matters: one correct trade-off makes the same planner reach an acceptable plan up to 4.6x as","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"},{"text":"The authors report that current agents admitted at most 32% of acceptable plans, while one correct trade-off admitted over 52%; in travel, showing plans raised F1 by 9.7 points over asking, with two model families.","excerpt":"Current agents, however, admit at most 32% of the acceptable plans, where that trade-off alone admits over 52%, because they ask about preferences instead of showing plans; in travel, showing plans raises F1 by 9.7 points over asking, with two model families.","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"},{"text":"The authors recommend testing a request against the user's reactions to concrete plans to identify trade-offs.","excerpt":"Planning agents should thus treat a request as a hypothesis to test against the user's reactions to concrete plans to recognize user trade-offs.","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.16466","checkedAt":"2026-11-13T12:50:09.345Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"Would showing plans improve identification of acceptable plans in settings beyond the travel comparison described in the abstract?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"arXiv cs.AI: How Should Planning Agents ReSHAPE Their Search Space?","summary":"Yilin Zhang and coauthors argue that planning agents should treat a user's request as a hypothesis, not a fixed specification. People may discover acceptable trade-offs only after seeing concrete plans, so an agent that asks about preferences without showing options can miss plans the user would accept. The authors formalize request revision as the ReSHAPE task: a revision is judged by whether its search space includes acceptable plans and a planner finds one. In ReSHAPE Planter, built from planning benchmarks with a human-validated simulated user, they analyze 5,800 agent-revised specifications in travel and shopping. They report that one correct trade-off lets the same planner find an acceptable plan up to 4.6x as often. Current agents include at most 32% of acceptable plans, compared with over 52% for that trade-off alone; in travel, showing plans raises F1 by 9.7 points over asking, with two model families. The authors recommend testing requests against users' reactions to concrete plans to recognize their trade-offs.","question":"Would showing plans improve identification of acceptable plans in settings beyond the travel comparison described in the abstract?","claims":[{"text":"Zhang and coauthors say users' initial requests can omit acceptable plans because people may not discover acceptable trade-offs until they see options.","excerpt":"the requests users write exclude plans they would gladly take, since they rarely discover the possible trade-offs they would accept until seeing the options.","support":"source-stated"},{"text":"The authors define ReSHAPE as judging whether a revised request's search space matches acceptable plans and a planner finds one there.","excerpt":"We formalize this step as the ReSHAPE task, which judges a revision by whether its search space matches the acceptable plans and a planner finds one there.","support":"source-stated"},{"text":"In their ReSHAPE Planter analysis of 5,800 agent-revised specifications in travel and shopping, one correct trade-off let the same planner reach an acceptable plan up to 4.6 times as often.","excerpt":"Using ReSHAPE Planter, which builds instances from planning benchmarks with a human-validated simulated user, we analyze 5,800 agent-revised specifications in travel and shopping. We find that reshaping matters: one correct trade-off makes the same planner reach an acceptable plan up to 4.6x as","support":"source-stated"},{"text":"The authors report that current agents admitted at most 32% of acceptable plans, while one correct trade-off admitted over 52%; in travel, showing plans raised F1 by 9.7 points over asking, with two model families.","excerpt":"Current agents, however, admit at most 32% of the acceptable plans, where that trade-off alone admits over 52%, because they ask about preferences instead of showing plans; in travel, showing plans raises F1 by 9.7 points over asking, with two model families.","support":"source-stated"},{"text":"The authors recommend testing a request against the user's reactions to concrete plans to identify trade-offs.","excerpt":"Planning agents should thus treat a request as a hypothesis to test against the user's reactions to concrete plans to recognize user trade-offs.","support":"source-stated"}],"limitations":["The abstract describes instances built from planning benchmarks with a human-validated simulated user, and reports analyses in travel and shopping; it does not establish how the results transfer to other settings or interactions with users outside the simulation.","The reported comparison of showing plans with asking is specifically described for travel."],"mediaId":"d0cca8ce6edff854edb06b5fcc76c1cced0990fdd6d9e6e063dbb63e92637f9d","correctionReason":null,"id":"dc794282b0c8a3665b595a22f79f507586cbafc23940cafeda2e249838675c2e","publicationId":"567d3fe1-5b6a-4473-8790-4fcad02a0d10","version":1,"sourceUrl":"https://arxiv.org/abs/2611.16466","sourceTitle":"How Should Planning Agents ReSHAPE Their Search Space?","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-12","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-13T07:40:02.745Z","createdAt":"2026-11-13T11:46:40.137Z","historical":false,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"d0cca8ce6edff854edb06b5fcc76c1cced0990fdd6d9e6e063dbb63e92637f9d","kind":"article","url":"https://arxiv.org/abs/2611.16466","sourceUrl":"https://arxiv.org/abs/2611.16466","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-13T11:34:23.770Z","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/2611.16466","checkedAt":"2026-11-13T12:50:09.345Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v7","mediaId":"d0cca8ce6edff854edb06b5fcc76c1cced0990fdd6d9e6e063dbb63e92637f9d","confidence":0.56,"provider":"typesafe","model":"jev-1.13.0","inputHash":"7918ae4c8d79d53ebaba0edbf9c6f64cd1d4b246ef98a3c65e28cd10848d4580","decidedAt":"2026-11-13T11:34:26.978Z"},"provenance":{"textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observationId":"509f528e248f54473d522082aba09fb92f71cbcb6805d37dfaf8a4b0f63a4f78","retrievedAt":"2026-11-13T11:34:23.236Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"9b6350df-a2f4-4ef2-89ca-9ebeb69aae6d"},"reviewer":{"model":"gpt-6-luna","runId":"1ebdb9cc-761f-4800-8e0e-d424e44b9651"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-13T11:46:40.137Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Zhang, Yilin","url":null,"kind":"unknown"},{"name":"Zhao, Xinran","url":null,"kind":"unknown"},{"name":"Chen, Tong","url":null,"kind":"unknown"},{"name":"Yuan, Ziming","url":null,"kind":"unknown"},{"name":"Wu, Tongshuang","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Zhang, Yilin","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhang, Yilin","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}},{"actor":{"name":"Zhao, Xinran","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhao, Xinran","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}},{"actor":{"name":"Chen, Tong","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Chen, Tong","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}},{"actor":{"name":"Yuan, Ziming","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Yuan, Ziming","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}},{"actor":{"name":"Wu, Tongshuang","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Wu, Tongshuang","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:4","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}}],"participationCoverage":{"total":5,"returned":5,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.16466","snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z","basis":"publisher-page-metadata","sourceLinks":44,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/dc794282b0c8a3665b595a22f79f507586cbafc23940cafeda2e249838675c2e/source"}},"canonicalUrl":"https://escapemeridian.com/readings/arxiv-cs-ai-how-should-planning-agents-reshape-their-search-dc794282","formats":{"html":"/readings/arxiv-cs-ai-how-should-planning-agents-reshape-their-search-dc794282","markdown":"/readings/arxiv-cs-ai-how-should-planning-agents-reshape-their-search-dc794282.md","json":"/readings/arxiv-cs-ai-how-should-planning-agents-reshape-their-search-dc794282.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"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"dc794282b0c8a3665b595a22f79f507586cbafc23940cafeda2e249838675c2e","revision":"567d3fe1-5b6a-4473-8790-4fcad02a0d10","revisionKind":"publication-id","title":"arXiv cs.AI: How Should Planning Agents ReSHAPE Their Search Space?","url":"https://escapemeridian.com/readings/arxiv-cs-ai-how-should-planning-agents-reshape-their-search-dc794282"},"source":{"type":"source","id":"https://arxiv.org/abs/2611.16466","revision":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","revisionKind":"sha256-source-text","title":"How Should Planning Agents ReSHAPE Their Search Space?","url":"https://arxiv.org/abs/2611.16466"},"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":"dc794282b0c8a3665b595a22f79f507586cbafc23940cafeda2e249838675c2e","title":"arXiv cs.AI: How Should Planning Agents ReSHAPE Their Search Space?","argument":"Yilin Zhang and coauthors argue that planning agents should treat a user's request as a hypothesis, not a fixed specification. People may discover acceptable trade-offs only after seeing concrete plans, so an agent that asks about preferences without showing options can miss plans the user would accept. The authors formalize request revision as the ReSHAPE task: a revision is judged by whether its search space includes acceptable plans and a planner finds one. In ReSHAPE Planter, built from planning benchmarks with a human-validated simulated user, they analyze 5,800 agent-revised specifications in travel and shopping. They report that one correct trade-off lets the same planner find an acceptable plan up to 4.6x as often. Current agents include at most 32% of acceptable plans, compared with over 52% for that trade-off alone; in travel, showing plans raises F1 by 9.7 points over asking, with two model families. The authors recommend testing requests against users' reactions to concrete plans to recognize their trade-offs.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The abstract describes instances built from planning benchmarks with a human-validated simulated user, and reports analyses in travel and shopping; it does not establish how the results transfer to other settings or interactions with users outside the simulation. The reported comparison of showing plans with asking is specifically described for travel.","speaker":{"name":"Zhang, Yilin, Zhao, Xinran, Chen, Tong, Yuan, Ziming, Wu, Tongshuang","roleAtSource":"Article author"},"source":{"title":"How Should Planning Agents ReSHAPE Their Search Space?","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.16466","publishedAt":"2026-11-12","dateBasis":"Date basis not provided."},"evidence":[{"text":"Zhang and coauthors say users' initial requests can omit acceptable plans because people may not discover acceptable trade-offs until they see options.","excerpt":"the requests users write exclude plans they would gladly take, since they rarely discover the possible trade-offs they would accept until seeing the options.","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"},{"text":"The authors define ReSHAPE as judging whether a revised request's search space matches acceptable plans and a planner finds one there.","excerpt":"We formalize this step as the ReSHAPE task, which judges a revision by whether its search space matches the acceptable plans and a planner finds one there.","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"},{"text":"In their ReSHAPE Planter analysis of 5,800 agent-revised specifications in travel and shopping, one correct trade-off let the same planner reach an acceptable plan up to 4.6 times as often.","excerpt":"Using ReSHAPE Planter, which builds instances from planning benchmarks with a human-validated simulated user, we analyze 5,800 agent-revised specifications in travel and shopping. We find that reshaping matters: one correct trade-off makes the same planner reach an acceptable plan up to 4.6x as","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"},{"text":"The authors report that current agents admitted at most 32% of acceptable plans, while one correct trade-off admitted over 52%; in travel, showing plans raised F1 by 9.7 points over asking, with two model families.","excerpt":"Current agents, however, admit at most 32% of the acceptable plans, where that trade-off alone admits over 52%, because they ask about preferences instead of showing plans; in travel, showing plans raises F1 by 9.7 points over asking, with two model families.","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"},{"text":"The authors recommend testing a request against the user's reactions to concrete plans to identify trade-offs.","excerpt":"Planning agents should thus treat a request as a hypothesis to test against the user's reactions to concrete plans to recognize user trade-offs.","url":"https://arxiv.org/abs/2611.16466","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.16466","checkedAt":"2026-11-13T12:50:09.345Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"Would showing plans improve identification of acceptable plans in settings beyond the travel comparison described in the abstract?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"arXiv cs.AI: How Should Planning Agents ReSHAPE Their Search Space?","summary":"Yilin Zhang and coauthors argue that planning agents should treat a user's request as a hypothesis, not a fixed specification. People may discover acceptable trade-offs only after seeing concrete plans, so an agent that asks about preferences without showing options can miss plans the user would accept. The authors formalize request revision as the ReSHAPE task: a revision is judged by whether its search space includes acceptable plans and a planner finds one. In ReSHAPE Planter, built from planning benchmarks with a human-validated simulated user, they analyze 5,800 agent-revised specifications in travel and shopping. They report that one correct trade-off lets the same planner find an acceptable plan up to 4.6x as often. Current agents include at most 32% of acceptable plans, compared with over 52% for that trade-off alone; in travel, showing plans raises F1 by 9.7 points over asking, with two model families. The authors recommend testing requests against users' reactions to concrete plans to recognize their trade-offs.","question":"Would showing plans improve identification of acceptable plans in settings beyond the travel comparison described in the abstract?","claims":[{"text":"Zhang and coauthors say users' initial requests can omit acceptable plans because people may not discover acceptable trade-offs until they see options.","excerpt":"the requests users write exclude plans they would gladly take, since they rarely discover the possible trade-offs they would accept until seeing the options.","support":"source-stated"},{"text":"The authors define ReSHAPE as judging whether a revised request's search space matches acceptable plans and a planner finds one there.","excerpt":"We formalize this step as the ReSHAPE task, which judges a revision by whether its search space matches the acceptable plans and a planner finds one there.","support":"source-stated"},{"text":"In their ReSHAPE Planter analysis of 5,800 agent-revised specifications in travel and shopping, one correct trade-off let the same planner reach an acceptable plan up to 4.6 times as often.","excerpt":"Using ReSHAPE Planter, which builds instances from planning benchmarks with a human-validated simulated user, we analyze 5,800 agent-revised specifications in travel and shopping. We find that reshaping matters: one correct trade-off makes the same planner reach an acceptable plan up to 4.6x as","support":"source-stated"},{"text":"The authors report that current agents admitted at most 32% of acceptable plans, while one correct trade-off admitted over 52%; in travel, showing plans raised F1 by 9.7 points over asking, with two model families.","excerpt":"Current agents, however, admit at most 32% of the acceptable plans, where that trade-off alone admits over 52%, because they ask about preferences instead of showing plans; in travel, showing plans raises F1 by 9.7 points over asking, with two model families.","support":"source-stated"},{"text":"The authors recommend testing a request against the user's reactions to concrete plans to identify trade-offs.","excerpt":"Planning agents should thus treat a request as a hypothesis to test against the user's reactions to concrete plans to recognize user trade-offs.","support":"source-stated"}],"limitations":["The abstract describes instances built from planning benchmarks with a human-validated simulated user, and reports analyses in travel and shopping; it does not establish how the results transfer to other settings or interactions with users outside the simulation.","The reported comparison of showing plans with asking is specifically described for travel."],"mediaId":"d0cca8ce6edff854edb06b5fcc76c1cced0990fdd6d9e6e063dbb63e92637f9d","correctionReason":null,"id":"dc794282b0c8a3665b595a22f79f507586cbafc23940cafeda2e249838675c2e","publicationId":"567d3fe1-5b6a-4473-8790-4fcad02a0d10","version":1,"sourceUrl":"https://arxiv.org/abs/2611.16466","sourceTitle":"How Should Planning Agents ReSHAPE Their Search Space?","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-12","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-13T07:40:02.745Z","createdAt":"2026-11-13T11:46:40.137Z","historical":false,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"d0cca8ce6edff854edb06b5fcc76c1cced0990fdd6d9e6e063dbb63e92637f9d","kind":"article","url":"https://arxiv.org/abs/2611.16466","sourceUrl":"https://arxiv.org/abs/2611.16466","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-13T11:34:23.770Z","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/2611.16466","checkedAt":"2026-11-13T12:50:09.345Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v7","mediaId":"d0cca8ce6edff854edb06b5fcc76c1cced0990fdd6d9e6e063dbb63e92637f9d","confidence":0.56,"provider":"typesafe","model":"jev-1.13.0","inputHash":"7918ae4c8d79d53ebaba0edbf9c6f64cd1d4b246ef98a3c65e28cd10848d4580","decidedAt":"2026-11-13T11:34:26.978Z"},"provenance":{"textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observationId":"509f528e248f54473d522082aba09fb92f71cbcb6805d37dfaf8a4b0f63a4f78","retrievedAt":"2026-11-13T11:34:23.236Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"9b6350df-a2f4-4ef2-89ca-9ebeb69aae6d"},"reviewer":{"model":"gpt-6-luna","runId":"1ebdb9cc-761f-4800-8e0e-d424e44b9651"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-13T11:46:40.137Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Zhang, Yilin","url":null,"kind":"unknown"},{"name":"Zhao, Xinran","url":null,"kind":"unknown"},{"name":"Chen, Tong","url":null,"kind":"unknown"},{"name":"Yuan, Ziming","url":null,"kind":"unknown"},{"name":"Wu, Tongshuang","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Zhang, Yilin","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhang, Yilin","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}},{"actor":{"name":"Zhao, Xinran","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhao, Xinran","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}},{"actor":{"name":"Chen, Tong","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Chen, Tong","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}},{"actor":{"name":"Yuan, Ziming","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Yuan, Ziming","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}},{"actor":{"name":"Wu, Tongshuang","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Wu, Tongshuang","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.16466","locator":"head/meta[name=author|citation_author]"},"assertionId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb:participation:4","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-13T12:50:09.345Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","sourceUrl":"https://arxiv.org/abs/2611.16466","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z"}}],"participationCoverage":{"total":5,"returned":5,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.16466","snapshotId":"f017f8d80e9e2efdde11291789a245967af04ae155a2913574227858d9eecadb","contentSha256":"8ae83078ac8acd3285bfc21efc401bf716b1d46b9a2e70052165dc034271bd2c","textHash":"ed2284e57d6729580f67354b102d9367c47f111ae222eb92f9b46c99339ca413","observedAt":"2026-11-13T12:50:09.345Z","basis":"publisher-page-metadata","sourceLinks":44,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/dc794282b0c8a3665b595a22f79f507586cbafc23940cafeda2e249838675c2e/source"}}}],"audits":[{"publicationId":"567d3fe1-5b6a-4473-8790-4fcad02a0d10","verdict":"uncertain","findings":[{"dimension":"attribution","status":"pass","explanation":"The reading attributes the argument, the ReSHAPE task, the experiments and the recommendation to Yilin Zhang and coauthors, matching the listed authors. The findings are framed as reported results, not established facts.","excerpt":"Authors:Yilin Zhang, Xinran Zhao, Tong Chen, Ziming Yuan, Tongshuang Wu"},{"dimension":"summary","status":"uncertain","explanation":"The reading is mostly faithful and keeps the paper's thesis and force. One passage shifts the criterion: the summary says a revision is judged by whether its search space 'includes' acceptable plans. The source says it 'matches' them, which also penalizes admitting unacceptable plans and is consistent with the F1 measure. Smallest repair: change 'includes' to 'matches'. Central terms 'search space' and 'F1' are not explained in plain language. These should get term-library tooltips, or be added to the library if missing, so non-specialists can follow.","excerpt":"judges a revision by whether its search space matches the acceptable plans and a planner finds one there"},{"dimension":"date","status":"pass","explanation":"originalPublished 2026-11-12 matches the arXiv v1 submission date. The feed date of 2026-11-13 is later and correctly not used.","excerpt":"[Submitted on 12 Nov 2026]"},{"dimension":"support","status":"pass","explanation":"The figures 5,800 specifications, up to 4.6x, at most 32% vs over 52%, and +9.7 F1 in travel with two model families all match the abstract, as does the recommendation. The limitations accurately note the simulated user and the travel-only scope of the comparison. The question follows from that scope rather than presuming an answer. No benefits or costs are invented.","excerpt":"in travel, showing plans raises F1 by 9.7 points over asking, with two model families"},{"dimension":"media","status":"pass","explanation":"The selected media is the verified direct link to the arXiv abstract page. That is the original source and the most direct access point.","excerpt":null}],"createdAt":"2026-11-13T11:46:40.137Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}