{
  "rubric_version": "PAPER-RUBRIC-1.0",
  "reviewer": "E",
  "paper_id": "ATT-0173",
  "paper_title": "Why Do Time Series Models Need Long Context Windows?",
  "paper_version": "arXiv 2606.01999v1",
  "version_date": "2026/06/01",
  "first_public_date": "2026/06/01",
  "assessment_date": "2026-11-08",
  "reading_record": {
    "completed_at_utc": "2026-11-08T02:30:20.266476+00:00",
    "pdf_pages": 34,
    "coverage": "逐页全文及附录文本阅读；关键公式或表格按需原页视觉核对。参考文献已检查关联定位，未逐篇读取全部前作。",
    "experimental_execution": false
  },
  "paper_type": [
    "方法或算法",
    "实证分析"
  ],
  "analysis_mode": "standard",
  "user_context": {
    "research_question": null,
    "prior_knowledge": null,
    "constraints": null
  },
  "legacy_review": {
    "read_before_new_review": true,
    "full_old_review_available": false,
    "history_counts_as_new_review": false,
    "record": {
      "candidate_id": "ATT-0173",
      "counts_as_review": false,
      "sidecar_files_supplied": false,
      "records": [
        {
          "source_record_id": "S5-173",
          "source_id": "S5",
          "source_library_file_id": "libfile_849759d66a808191813dbe9bd1605bb4",
          "source_filename": "已粘贴的文本(5).txt",
          "source_line": 199,
          "source_number": 173,
          "title": "Why Do Time Series Models Need Long Context Windows?",
          "source_url": "https://neurips.cc/virtual/2026/poster/149819",
          "raw_line": "| P173 | [Why Do Time Series Models Need Long Context Windows?](https://neurips.cc/virtual/2026/poster/149819) | 0 | 待处理 | 尚未评审；引用较低（0 次，低于 10 次门槛） |",
          "raw_cells": [
            "P173",
            "[Why Do Time Series Models Need Long Context Windows?](https://neurips.cc/virtual/2026/poster/149819)",
            "0",
            "待处理",
            "尚未评审；引用较低（0 次，低于 10 次门槛）"
          ],
          "normalized_title": "whydotimeseriesmodelsneedlongcontextwindows",
          "normalized_url": "neurips.cc/virtual/2026/poster/149819",
          "record_type": "legacy_unranked",
          "citation_verbatim": "0",
          "legacy_status_verbatim": "待处理",
          "legacy_unranked_reason_verbatim": "尚未评审；引用较低（0 次，低于 10 次门槛）",
          "counts_as_independent_review": false
        }
      ]
    },
    "handling": "旧完整评语及评分sidecar未提供；历史条目仅作为待核线索，不继承旧分数。"
  },
  "claims": [
    {
      "claim_id": "CL01",
      "claim": "过程后验不确定性可使全局模型需要超过真实记忆阶数的输入。",
      "claim_type": "理论/解释",
      "conditions": "时不变随机过程族、加性独立噪声、MSE、假设3.1",
      "support_status": "充分支持（必要条件范围内）",
      "source_ids": [
        "S01"
      ]
    },
    {
      "claim_id": "CL02",
      "claim": "真实基础模型利用非邻接历史支持GPI解释。",
      "claim_type": "机制",
      "conditions": "四个真实数据集、固定两周间隔",
      "support_status": "部分支持",
      "source_ids": [
        "S01"
      ]
    },
    {
      "claim_id": "CL03",
      "claim": "解耦并缓存历史可改善重复预测成本效果。",
      "claim_type": "效率",
      "conditions": "历史嵌入可复用，允许特定准确率预算",
      "support_status": "部分支持",
      "source_ids": [
        "S01"
      ]
    }
  ],
  "scores": [
    {
      "dimension_id": "A1",
      "dimension_name": "问题重要性",
      "score": 85,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "长窗口成本与全局预测识别困难有明确领域意义。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p1—3"
    },
    {
      "dimension_id": "A2",
      "dimension_name": "原创性",
      "score": 75,
      "status": "tentative",
      "confidence": "中",
      "judgment_range": [
        65,
        85
      ],
      "rationale": "把贝叶斯分解转为GPI/CF设计框架；前作独立核对有限。",
      "missing_information": "把贝叶斯分解转为GPI/CF设计框架；前作独立核对有限。",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p3,9,22—23"
    },
    {
      "dimension_id": "A3",
      "dimension_name": "实质推进程度",
      "score": 75,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "合成可识别性与真实数据解耦收益清楚，未建新基础模型。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p5—8"
    },
    {
      "dimension_id": "A4",
      "dimension_name": "知识与解释增量",
      "score": 80,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "给出过程后验方差解释，真实数据仍有周期与领域混杂。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p3—7,13—17"
    },
    {
      "dimension_id": "B1",
      "dimension_name": "问题定义与评价有效性",
      "score": 80,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "MSE理论与实验相符；真实数据MAE不直接等价MSE分解。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p3—8"
    },
    {
      "dimension_id": "B2",
      "dimension_name": "技术与论证正确性",
      "score": 75,
      "status": "tentative",
      "confidence": "中",
      "judgment_range": [
        65,
        85
      ],
      "rationale": "主方差推导成立；W<P论述简略且噪声尺度标注冲突。",
      "missing_information": "主方差推导成立；W<P论述简略且噪声尺度标注冲突。",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p13—18"
    },
    {
      "dimension_id": "B3",
      "dimension_name": "数据与评测流程完整性",
      "score": 80,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "时间和序列双轴划分、独立序列验证集可追溯。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p19"
    },
    {
      "dimension_id": "B4",
      "dimension_name": "比较对象与比较公平性",
      "score": 65,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "对照架构丰富；基础模型对专门模型并非等数据等预算。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p6—8,20—21"
    },
    {
      "dimension_id": "C1",
      "dimension_name": "核心主张的证据充分度",
      "score": 75,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "必要条件、受控NAR及真实数据互补；并非所有效率条件均验证。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p5—8,24—32"
    },
    {
      "dimension_id": "C2",
      "dimension_name": "机制归因与竞争解释区分度",
      "score": 65,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "隔离历史与编码器替换有针对性，周期和参数增加未完全排除。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p4,17,24"
    },
    {
      "dimension_id": "C3",
      "dimension_name": "结果稳定性与不确定性处理",
      "score": 75,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "多数三运行、线性五运行并报告标准差，随机源有所交代。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p16,19,24—25"
    },
    {
      "dimension_id": "C4",
      "dimension_name": "适用范围与边界验证",
      "score": 70,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "多域与阶数覆盖充分，快速漂移和完整基础模型仍留待研究。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p9,24—25"
    },
    {
      "dimension_id": "D1",
      "dimension_name": "过程透明度",
      "score": 75,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "配置划分成本细致，但噪声方差和标准差不一致影响重建。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p18—22"
    },
    {
      "dimension_id": "D2",
      "dimension_name": "研究资源可获得性",
      "score": 50,
      "status": "tentative",
      "confidence": "中",
      "judgment_range": [
        35,
        75
      ],
      "rationale": "PDF及数据入口明确；CER需申请，核心代码本次未验证。",
      "missing_information": "PDF及数据入口明确；CER需申请，核心代码本次未验证。",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p18,21"
    },
    {
      "dimension_id": "D3",
      "dimension_name": "主张克制与报告一致性",
      "score": 65,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "限制有披露，但真实数据GPI归因与普遍Pareto用语偏强。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p7—9,26"
    },
    {
      "dimension_id": "D4",
      "dimension_name": "表达清晰度",
      "score": 80,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "主线清楚，附录充分；个别符号和图例文字不一致。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p3—9,18"
    },
    {
      "dimension_id": "E1",
      "dimension_name": "复杂度合理性",
      "score": 80,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "长历史编码与短窗口预测职责清楚，方案可分离。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p7—8,20"
    },
    {
      "dimension_id": "E2",
      "dimension_name": "资源效率与成本收益",
      "score": 70,
      "status": "tentative",
      "confidence": "中",
      "judgment_range": [
        55,
        80
      ],
      "rationale": "CPU/GPU/FLOPs/显存均测量，固定缓存和完整训练成本须限定。",
      "missing_information": "CPU/GPU/FLOPs/显存均测量，固定缓存和完整训练成本须限定。",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p21—22,26—32"
    },
    {
      "dimension_id": "E3",
      "dimension_name": "可迁移与可复用程度",
      "score": 85,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "识别与预测分工可迁移到多种全局架构。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p7,20—23"
    },
    {
      "dimension_id": "E4",
      "dimension_name": "后续研究启发性",
      "score": 85,
      "status": "rated",
      "confidence": "高",
      "judgment_range": null,
      "rationale": "形成过程异质性、窗口长度与缓存刷新之间的具体问题。",
      "missing_information": "",
      "claim_ids": [],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p5—9"
    },
    {
      "dimension_id": "F1",
      "dimension_name": "与当前问题的相关性",
      "score": null,
      "status": "unknown",
      "confidence": null,
      "judgment_range": null,
      "rationale": "未给定用户正在解决的具体预测或异常检测问题。",
      "missing_information": "个人研究背景缺失；不以论文质量代填。",
      "claim_ids": [],
      "source_ids": [],
      "location": "任务上下文（非论文证据）"
    },
    {
      "dimension_id": "F2",
      "dimension_name": "相对于已有知识的学习收益",
      "score": null,
      "status": "unknown",
      "confidence": null,
      "judgment_range": null,
      "rationale": "未给定用户已掌握的理论、算法或已读论文。",
      "missing_information": "个人研究背景缺失；不以论文质量代填。",
      "claim_ids": [],
      "source_ids": [],
      "location": "任务上下文（非论文证据）"
    },
    {
      "dimension_id": "F3",
      "dimension_name": "当前使用可行性",
      "score": null,
      "status": "unknown",
      "confidence": null,
      "judgment_range": null,
      "rationale": "未给定用户用途、算力、数据与许可约束。",
      "missing_information": "个人研究背景缺失；不以论文质量代填。",
      "claim_ids": [],
      "source_ids": [],
      "location": "任务上下文（非论文证据）"
    },
    {
      "dimension_id": "F4",
      "dimension_name": "当前决策信息价值",
      "score": null,
      "status": "unknown",
      "confidence": null,
      "judgment_range": null,
      "rationale": "未给定当前待决策的具体方法或研究路线。",
      "missing_information": "个人研究背景缺失；不以论文质量代填。",
      "claim_ids": [],
      "source_ids": [],
      "location": "任务上下文（非论文证据）"
    }
  ],
  "risks": [
    {
      "risk_id": "R01",
      "severity": "重要问题",
      "certainty": "已确认报告不一致；实现原因未知",
      "description": "公式27将0.2称方差，最小MSE曲线却约0.04；不能由论文唯一复建合成噪声尺度。",
      "affected_claim_ids": [
        "CL01"
      ],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p5,18,24—25"
    },
    {
      "risk_id": "R02",
      "severity": "重要问题",
      "certainty": "证据局限",
      "description": "非邻接历史仍可能携带季节信息，基础模型与专门模型也非等预算，真实GPI归因不能独占。",
      "affected_claim_ids": [
        "CL02"
      ],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p4,6—7"
    },
    {
      "risk_id": "R03",
      "severity": "重要问题",
      "certainty": "适用条件明确",
      "description": "效率结果依赖历史固定复用；首轮、刷新与快速漂移成本未充分验证。",
      "affected_claim_ids": [
        "CL03"
      ],
      "source_ids": [
        "S01"
      ],
      "location": "PDF p8—9,21—22,26"
    }
  ],
  "aggregation": {
    "weights": null,
    "weighted_score_observed": null,
    "weighted_coverage": null,
    "tentative_weight_share": null,
    "numeric_dimensions": 20,
    "applicable_dimensions": 24,
    "unweighted_dimension_coverage": 0.8333333333333334,
    "notes": "保留24维原始分数，不计算论文总分或跨论文排序；F组未知。"
  },
  "reading_recommendation": "重点阅读，尤其用于长上下文机制与推理成本分析",
  "recommended_sections": [
    "p3—5 公式5与定理3.2",
    "p21—22 附录F—G及缓存条件"
  ],
  "sources": [
    {
      "source_id": "S01",
      "source_type": "paper_pdf",
      "title": "Why Do Time Series Models Need Long Context Windows?",
      "version_or_date": "2606.01999v1",
      "location": "/workspace/scratch/ac01e91af073/candidate_recovery/priority_packet_20261108/papers/ATT-0173/ATT-0173_paper.pdf",
      "sha256": "9226e890eaf0c0306caec014475d6c123b4185d6727e08da42b26b453351a0c8",
      "access_status": "已读取完整PDF文本及附录；本地版本，不推定等同会议终稿"
    },
    {
      "source_id": "S02",
      "source_type": "legacy_history",
      "location": "/workspace/scratch/ac01e91af073/candidate_recovery/priority_packet_20261108/papers/ATT-0173/legacy_reference_only.json",
      "access_status": "已先读；无旧完整评语"
    }
  ],
  "independence": "未读取同批其他评审的新意见。",
  "narrative_chinese_characters": 1754
}