arXiv cs.CL· Jonathan Zheng, Zirui Shao, Alan Ritter, Wei Xu·· 13 小时前AI 评分38
PARALLELEVENTS 基准与 SYNAPSE 框架：用模拟未来事件评估和训练 LLM 知识更新
Parallel Universes for Temporal Evaluation and Knowledge Updating in LLMs
AI 导读
研究者提出 PARALLELEVENTS 基准，通过模拟虚构但合理的未来事件轨迹，在无数据污染、不与既有知识冲突的前提下评估 LLM 的知识插入能力。配套的 SYNAPSE 训练框架利用模型生成数据，经指令微调将新知识写入模型参数，无需昂贵的人工标注数据。该方法在问答准确率上较此前方法提升 14.23 分，论文已被 NeurIPS 2026 主会接收。
正文 · 原文
Abstract:Large language models (LLMs) rely on static pretraining corpora, resulting in outdated knowledge over time. Existing approaches for evaluating knowledge edits either suffer from rapid contamination or rely on counterfactual edits that conflict with rigid existing knowledge. In this work, we propose a simulation-driven approach for evaluating and training LLMs on knowledge insertions. We introduce PARALLELEVENTS, a benchmark that simulates coherent trajectories of fictional yet realistic future events, enabling controlled evaluation without data contamination or conflicts with existing knowledge. We further propose SYNAPSE, a training framework that leverages model-generated data to integrate new knowledge into model parameters via instruction tuning. This synthetic pipeline enables scalable knowledge integration without costly human-curated data. SYNAPSE improves QA accuracy by 14.23 points over prior methods, indicating that simulation-based synthetic training yields robust, coherent knowledge insertion.
| Comments: | accepted to NeurIPS 2026 Main Conference, 10 pages |
| Subjects: | Computation and Language (cs.CL); Machine Learning (cs.LG) |
| Cite as: | arXiv:2609.00184 [cs.CL] |
| (or arXiv:2609.00184v3 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2609.00184 arXiv-issued DOI via DataCite |
Submission history
From: Jonathan Zheng [view email]
[v1]
Mon, 31 Aug 2026 18:06:35 UTC (6,829 KB)
[v2]
Fri, 4 Sep 2026 17:29:17 UTC (6,829 KB)
[v3]
Mon, 26 Oct 2026 23:01:49 UTC (5,896 KB)
来源：arXiv cs.CL · arxiv.org