{"record": {"aliases": ["TREK: A Travel Reasoning and Evaluation Kit for LLM Agents in Complex Trip Planning"], "artifacts": [{"id": "arxiv:2607.26977", "kind": "paper", "url": "https://arxiv.org/abs/2607.26977"}, {"id": "gh:tonyqjh/trek-a-travel-reasoning-and-evaluation-kit-for-llm-agents-in-complex-trip-planning", "kind": "repo", "url": "https://github.com/TonyQJH/TREK-A-Travel-Reasoning-and-Evaluation-Kit-for-LLM-Agents-in-Complex-Trip-Planning"}], "authors": ["Jinhu Qi", "Wentao Zhang", "Siu Man Ng", "Feiyang Xu", "Yanyu Chen", "Yaoman Li", "Irwin King"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.26977", "source_url": "https://arxiv.org/abs/2607.26977"}, "categories": ["General AI", "Language & Knowledge", "Planning", "Reasoning"], "description": {"en": "TREK evaluates agents on synthesizing feasible travel itineraries under interacting entity, route, time, budget, and persona constraints."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.26977", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.26977", "title": "TREK"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Chinese University of Hong Kong", "Macao Polytechnic University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2607.26977", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2607.26977", "source_url": "https://openalex.org/W7171789461"}, "key": "claire-radar:2607.26977", "modality": null, "name": "TREK", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-09-25T04:40:22Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_trek-a-travel-reasoning-and-evaluation-kit_17e93ec1", "origin_source": "arxiv", "recognition_confidence": "0.7", "record_sha256": "ed321a5f064af8eb2a7710fa096ac0bfb193cc7cc98645b288cfb8ea23bea407", "relation": "introduces", "review_model": "codex-cli-default", "review_state": "ai-reviewed", "reviewed_at": "2026-09-24T17:32:36.949824Z", "source_url": "https://arxiv.org/abs/2607.26977"}, "publisher": null, "released": "2026-07-29", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.26977", "source_url": "https://arxiv.org/abs/2607.26977"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-26977", "source": "claire_radar", "source_benchmark_id": "2607.26977", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["TREK: A Travel Reasoning and Evaluation Kit for LLM Agents in Complex Trip Planning"], "annotation": "Machine Generated", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The large task set, deterministic feasibility checks, and reproducible agent environment should generate interest among agent-evaluation researchers.", "score": 72}, "capabilities": ["Reasoning", "Planning"], "confidence": "Medium", "construction": "Original Synthetic", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-24T17:32:36.949824Z", "inputHash": "31f34703d52685b46e7380761d86edda57c436b6eaec712823881ffcf7bd73c3", "model": "codex-cli-default"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "TREK: A Travel Reasoning and Evaluation Kit for LLM Agents in Complex Trip Planning", "canonicalNameSource": "paper_title", "decisionReason": "The supplied evidence identifies a named benchmark, a defined synthesis task, deterministic rule-based scoring, and a public release containing the dataset, sandbox, evaluator, and agent code.", "evaluationEvidence": {"reuse": "We release the dataset, tool sandbox, deterministic evaluator, and agent code as a fully reproducible benchmark.", "scoring": "Every task is scored by a fully deterministic, rule-based evaluator with no LLM judge", "task": "a benchmark for feasible itinerary synthesis"}, "model": "codex-cli-default", "reviewedAt": "2026-09-24T17:32:36.949824Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "TREK evaluates agents on synthesizing feasible travel itineraries under interacting entity, route, time, budget, and persona constraints.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce TREK (Travel Reasoning and Evaluation Kit), a benchmark for feasible itinerary synthesis: producing a single plan that is jointly constraint-correct, hallucination-free, spatio-temporally executable, budget-valid, and responsive to the traveler's unstated persona needs."}, "familyId": "bmf_e86d910a01ee", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_trek-a-travel-reasoning-and-evaluation-kit_17e93ec1", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/TonyQJH/TREK-A-Travel-Reasoning-and-Evaluation-Kit-for-LLM-Agents-in-Complex-Trip-Planning", "data": null, "paper": "https://arxiv.org/abs/2607.26977", "pdf": "https://arxiv.org/pdf/2607.26977", "project": null, "report": "https://arxiv.org/abs/2607.26977"}, "metrics": [], "motivation": "Travel planning is a demanding stress test for tool-using LLM agents: a usable itinerary is a single artifact that must be right along many axes at once - every flight, hotel, and attraction must exist and be bookable, the days must be physically traversable, the total must clear a budget, and the plan must serve a traveler whose needs are only partly stated.", "name": "TREK", "oneLine": "TREK evaluates agents on synthesizing feasible travel itineraries under interacting entity, route, time, budget, and persona constraints.", "paperTitle": "TREK: A Travel Reasoning and Evaluation Kit for LLM Agents in Complex Trip Planning", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.26977", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "publishers": [{"name": "TREK contributors", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://github.com/TonyQJH/TREK-A-Travel-Reasoning-and-Evaluation-Kit-for-LLM-Agents-in-Complex-Trip-Planning"}], "readiness": "Runnable", "recognitionConfidence": 0.7, "relation": "introduces", "releasedAt": "2026-07-29", "reviewLifecycle": {"state": "pending"}, "source": {"authors": ["Jinhu Qi", "Wentao Zhang", "Siu Man Ng", "Feiyang Xu", "Yanyu Chen", "Yaoman Li", "Irwin King"], "categories": ["cs.CL"], "id": "2607.26977", "title": "TREK: A Travel Reasoning and Evaluation Kit for LLM Agents in Complex Trip Planning", "type": "arxiv", "url": "https://arxiv.org/abs/2607.26977"}, "sourceUpdatedAt": "2026-08-10T02:07:10Z", "topics": ["Reasoning"], "venueAttempts": [], "whyItMatters": "Its deterministic evaluator and validated synthetic environment make end-to-end itinerary feasibility reproducible and auditable."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
