至少在ICLR 2026的投稿中发现了50篇幻觉引文。
Over fifty new hallucinations in ICLR 2026 submissions

原始链接: https://gptzero.me/news/iclr-2026/

本文详细记录了对53篇研究论文引用的扫描结果,揭示了大型语言模型(LLM)生成的参考文献存在显著问题。使用引文验证工具的分析发现,不准确现象普遍存在。 超过60%的论文列表存在问题,范围从完全捏造的作者和标题到不正确的细节,如出版年份或网址。许多引用*部分*匹配现有论文,但作者列表或关键元数据有重大更改。在一些情况下,提供的引用指向不相关的文章,甚至不存在的来源。 这些错误并非随机的;有些论文完全捏造了作者列表,而另一些论文则正确引用了论文,但修改了细节。这表明LLM并非仅仅是记错,而是主动构建听起来合理但错误的引用。这些“幻觉”的频率凸显了在使用LLM进行学术研究和文献综述时的一个关键可靠性问题。与每篇论文相关的评分范围为0.5到8.0,表明受影响的研究领域广泛。

最近一篇Hacker News上的帖子指出了一种令人担忧的趋势:在提交给2026年国际表示学习会议(ICLR)的论文中,出现了广泛的幻觉引文现象。对20,000篇论文中的300篇初步扫描已经发现了至少50条伪造的引文,估计数量可能还会更多。 讨论的中心是这些错误是否完全归因于大型语言模型(LLM),以及建立LLM出现之前的论文的基线错误率的必要性。评论员强调了同行评审期间彻底检查引文的重要性——这个过程本身就容易出现人为错误,正如最近一期期刊的错误所证明。 人们还担心人工智能生成的不准确信息可能带来的法律影响,一位用户建议可能存在责任问题,甚至在法律等领域完全禁止使用人工智能。这场对话凸显了人们对将这些错误定义为任何其他词语(例如“谎言”或“捏造”)的日益沮丧,并强调了人类作者验证人工智能辅助工作的责任。
相关文章

原文
TitleAverage Review RatingPaper LinkCitation Check Scan LinkExample of Verified HallucinationCommentTamperTok: Forensics-Driven Tokenized Autoregressive Framework for Image Tampering Localization8.0TamperTok: Forensics-Driven Tokenized Autoregressive Framework for Image Tampering Localization | OpenReviewhttps://app.gptzero.me/documents/4645494f-70eb-40bb-aea7-0007e13f7179/shareChong Zou, Zhipeng Wang, Ziyu Li, Nan Wu, Yuling Cai, Shan Shi, Jiawei Wei, Xia Sun, Jian Wang, and Yizhou Wang. Segment everything everywhere all at once. In Advances in Neural Information Processing Systems (NeurIPS), volume 36, 2023.This paper exists, but all authors are wrong.MixtureVitae: Open Web-Scale Pretraining Dataset With High Quality Instruction and Reasoning Data Built from Permissive Text Sources8.0MixtureVitae: Open Web-Scale Pretraining Dataset With High Quality Instruction and Reasoning Data Built from Permissive Text Sources | OpenReviewhttps://app.gptzero.me/documents/bfd10666-ea2d-454c-9ab2-75faa8b84281/shareDan Hendrycks, Collin Burns, Steven Basart, Andy Critch, Jerry Li, Dawn Ippolito, Aina Lapedriza, Florian Tramer, Rylan Macfarlane, Eric Jiang, et al. Measuring massive multitask language understanding. In Proceedings of the International Conference on Learning Representations (ICLR), 2021.The paper and first 3 authors match. The last 7 authors are not on the paper, and some of them do not existCatch-Only-One: Non-Transferable Examples for Model-Specific Authorization6.0Catch-Only-One: Non-Transferable Examples for Model-Specific Authorization | OpenReviewhttps://app.gptzero.me/documents/9afb1d51-c5c8-48f2-9b75-250d95062521/shareDinghuai Zhang, Yang Song, Inderjit Dhillon, and Eric Xing. Defense against adversarial attacks using spectral regularization. In International Conference on Learning Representations (ICLR), 2020.No MatchOrtSAE: Orthogonal Sparse Autoencoders Uncover Atomic Features6.0OrtSAE: Orthogonal Sparse Autoencoders Uncover Atomic Features | OpenReviewhttps://app.gptzero.me/documents/e3f155d7-067a-4720-adf8-65dc9dc714b9/shareRobert Huben, Logan Riggs, Aidan Ewart, Hoagy Cunningham, and Lee Sharkey. Sparse autoencoders can interpret randomly initialized transformers, 2025. URL https://arxiv.org/ abs/2501.17727.This paper exists, but all authors are wrong.Principled Policy Optimization for LLMs via Self-Normalized Importance Sampling5.0Principled Policy Optimization for LLMs via Self-Normalized Importance Sampling | OpenReviewhttps://app.gptzero.me/documents/54c8aa45-c97d-48fc-b9d0-d491d54df8d3/shareDavid Rein, Stas Gaskin, Lajanugen Logeswaran, Adva Wolf, Oded teht sun, Jackson H. He, Divyansh Kaushik, Chitta Baral, Yair Carmon, Vered Shwartz, Sang-Woo Lee, Yoav Goldberg, C. J. H. un, Swaroop Mishra, and Daniel Khashabi. Gpqa: A graduate-level google-proof q\&a benchmark, 2023All authors except the first are fabricated.PDMBench: A Standardized Platform for Predictive Maintenance Research4.5PDMBench: A Standardized Platform for Predictive Maintenance Research | OpenReviewhttps://app.gptzero.me/documents/5c55afe7-1689-480d-ac44-9502dc0f9229/share Andrew Chen, Andy Chow, Aaron Davidson, Arjun DCunha, Ali Ghodsi, Sue Ann Hong, Andy Konwinski, Clemens Mewald, Siddharth Murching, Tomas Nykodym, et al. Mlflow: A platform for managing the machine learning lifecycle. In Proceedings of the Fourth International Workshop on Data Management for End-to-End Machine Learning, pp. 1-4. ACM, 2018. Authors and conference match this paper, but title is somewhat different and the year is wrong.IMPQ: Interaction-Aware Layerwise Mixed Precision Quantization for LLMs4.5IMPQ: Interaction-Aware Layerwise Mixed Precision Quantization for LLMs | OpenReviewhttps://app.gptzero.me/documents/5461eefd-891e-4100-ba1c-e5419af520c0/shareChen Zhu et al. A survey on efficient deployment of large language models. arXiv preprint arXiv:2307.03744, 2023.The arXiv ID is real, but the paper has different authors and a different title.C3-OWD: A Curriculum Cross-modal Contrastive Learning Framework for Open-World Detection4.5C3-OWD: A Curriculum Cross-modal Contrastive Learning Framework for Open-World Detection | OpenReviewhttps://app.gptzero.me/documents/c07521cd-2757-40a2-8dc1-41382d7eb11b/shareK. Marino, R. Salakhutdinov, and A. Gupta. Fine-grained image classification with learnable semantic parts. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 4500-4509, 2019.Authors and subject match this paperTopoMHC: Sequence–Topology Fusion for MHC Binding4.5TopoMHC: Sequence–Topology Fusion for MHC Binding | OpenReviewhttps://app.gptzero.me/documents/8da4f86c-00d8-4d73-81dd-c168c0bfdf4e/shareYuchen Han, Yohan Kim, Dalibor Petrovic, Alessandro Sette, Morten Nielsen, and Bjoern Peters. Deepligand: a deep learning framework for peptide-mhc binding prediction. Bioinformatics, 39 (1):btac834, 2023. doi: 10.1093/bioinformatics/btac834.No MatchCan Text-to-Video Models Generate Realistic Human Motion?4.5Can Text-to-Video Models Generate Realistic Human Motion? | OpenReviewhttps://app.gptzero.me/documents/f52aad2d-2253-44bf-80ba-8e8668df650f/shareYugandhar Balaji, Jianwei Yang, Zhen Xu, Menglei Chai, Zhoutong Xu, Ersin Yumer, Greg Shakhnarovich, and Deva Ramanan. Conditional gan with discriminative filter generation for text-to-video synthesis. In Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI), pp. 2155-2161, July 2019. doi: 10.24963/ijcai.2019/276.This paper exists, but the authors and page numbers are wrong.GRF-LLM: Environment-Aware Wireless Channel Modeling via LLM-Guided 3D Gaussians4.0GRF-LLM: Environment-Aware Wireless Channel Modeling via LLM-Guided 3D Gaussians | OpenReviewhttps://app.gptzero.me/documents/c3e66b9c-20b4-4c50-b881-e40aba2a514f/shareJunting Chen, Yong Zeng, and Rui Zhang. Rfcanvas: A radio frequency canvas for wireless network design. In IEEE International Conference on Communications, pp. 1-6, 2024.
Title partially matches this article.Listwise Generalized Preference Optimization with Process-aware Signals for LLM Reasoning4.0Listwise Generalized Preference Optimization with Process-aware Signals for LLM Reasoning | OpenReviewhttps://app.gptzero.me/documents/bbeecf1c-189a-4311-999b-617aab686ea9/shareKaixuan Zhou, Jiaqi Liu, Yiding Wang, and James Zou. Generalized direct preference optimization. arXiv preprint arXiv:2402.05015, 2024.No MatchIUT-Plug: A Plug-in tool for Interleaved Image-Text Generation4.0IUT-Plug: A Plug-in tool for Interleaved Image-Text Generation | OpenReviewhttps://app.gptzero.me/documents/0f12d2fc-403b-4859-8d00-f75fd9f56e39/shareYash Goyal, Anamay Mohapatra, Nihar Kwatra, and Pawan Goyal. A benchmark for compositional text-to-image synthesis. In Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 1), 2021.This paper exists, but the authors are all wrong.Resolving the Security-Auditability Dilemma with Auditable Latent Chain-of-Thought4.0Resolving the Security-Auditability Dilemma with Auditable Latent Chain-of-Thought | OpenReviewhttps://app.gptzero.me/documents/5cee5c3a-5e75-4063-a054-1e934a071705/shareYixiang Ma, Ziyi Liu, Zhaoyu Wang, Zhaofeng Xu, Yitao Wang, and Yang Liu. Safechain: A framework for securely executing complex commands using large language models. arXiv preprint arXiv:2402.16521, 2024a.No match; although this paper is closely related.ThinkGeo: Evaluating Tool-Augmented Agents for Remote Sensing Tasks4.0ThinkGeo: Evaluating Tool-Augmented Agents for Remote Sensing Tasks | OpenReviewhttps://app.gptzero.me/documents/f3441445-5401-48e9-9617-09a635992ff9/shareYunzhu Yang, Shuang Li, and Jiajun Wu. MM-ReAct: Prompting chatgpt to multi-modal chain-ofthought reasoning. arXiv preprint arXiv:2401.04740, 2024.No MatchTaming the Judge: Deconflicting AI Feedback for Stable Reinforcement Learning3.5Taming the Judge: Deconflicting AI Feedback for Stable Reinforcement Learning | OpenReviewhttps://app.gptzero.me/documents/80c64df2-eee6-41aa-90cc-3f835b128747/shareChenglong Wang, Yang Liu, Zhihong Xu, Ruochen Zhang, Jiahao Wu, Tao Luo, Jingang Li, Xunliang Liu, Weiran Qi, Yujiu Yang, et al. Gram-r ${ }^{8}$ : Self-training generative foundation reward models for reward reasoning. arXiv preprint arXiv:2509.02492, 2025b.All authors except the first are fabricated and the title is altered.DANCE-ST: Why Trustworthy AI Needs Constraint Guidance, Not Constraint Penalties3.5DANCE-ST: Why Trustworthy AI Needs Constraint Guidance, Not Constraint Penalties | OpenReviewhttps://app.gptzero.me/documents/3ebd71b4-560d-4fa3-a0d3-ed2fa13c519f/shareSardar Asif, Saad Ghayas, Waqar Ahmad, and Faisal Aadil. Atcn: an attention-based temporal convolutional network for remaining useful life prediction. The Journal of Supercomputing, 78(1): $1-19,2022$.Two papers with similar titles exist here and here, but the authors, journal, and date do not match.Federated Hierarchical Anti-Forgetting Framework for Class-Incremental Learning with Large Pre-Trained Models3.33Federated Hierarchical Anti-Forgetting Framework for Class-Incremental Learning with Large Pre-Trained Models | OpenReviewhttps://app.gptzero.me/documents/ae10437b-c65b-455b-ad22-918742a5ed82/shareArslan Chaudhry, Arun Mallya, and Abhinav Srivastava. Fedclassil: A benchmark for classincremental federated learning. In NeurIPS, 2023.No MatchChain-of-Influence: Tracing Interdependencies Across Time and Features in Clinical Predictive Modeling3.33Chain-of-Influence: Tracing Interdependencies Across Time and Features in Clinical Predictive Modeling | OpenReviewhttps://app.gptzero.me/documents/dff2c063-6986-4241-8c20-4327a39d4d4b/shareIshita et al. Bardhan. Icu length-of-stay prediction with interaction-based explanations. Journal of Biomedical Informatics, 144:104490, 2024.No MatchTRACEALIGN - Tracing the Drift: Attributing Alignment Failures to Training-Time Belief Sources in LLMs3.33TRACEALIGN - Tracing the Drift: Attributing Alignment Failures to Training-Time Belief Sources in LLMs | OpenReviewhttps://app.gptzero.me/documents/4b379aba-8d8a-427b-ac67-d13af5eda8c9/shareLisa Feldman Barrett. Emotions are constructed: How brains make meaning. Current Directions in Psychological Science, 25(6):403-408, 2016.This article is similar, but the title, and metadata are different.MEMORIA: A Large Language Model, Instruction Data and Evaluation Benchmark for Intangible Cultural Heritage3.33MEMORIA: A Large Language Model, Instruction Data and Evaluation Benchmark for Intangible Cultural Heritage | OpenReviewhttps://app.gptzero.me/documents/956129a3-11ee-4503-92e3-3ed5db12d2d6/shareYang Cao, Rosa Martinez, and Sarah Thompson. Preserving indigenous languages through neural language models: Challenges and opportunities. Computational Linguistics, 49(3):567-592, 2023.No MatchReflexion: Language Models that Think Twice for Internalized Self-Correction3.2Reflexion: Language Models that Think Twice for Internalized Self-Correction | OpenReviewhttps://app.gptzero.me/documents/45f2f68d-df09-4bbf-8513-588fe24f26fa/shareGuang-He Xiao, Haolin Wang, and Yong-Feng Zhang. Rethinking uncertainty in llms: A case study on a fact-checking benchmark. arXiv preprint arXiv:2305.11382, 2023.No MatchECAM: Enhancing Causal Reasoning in Foundation Models with Endogenous Causal Attention Mechanism3.0ECAM: Enhancing Causal Reasoning in Foundation Models with Endogenous Causal Attention Mechanism | OpenReviewhttps://app.gptzero.me/documents/d99a5552-38e0-459b-8746-4e64069b0640/shareAtticus Geiger, Zhengxuan Wu, Yonatan Rozner, Mirac Suzgun Naveh, Anna Nagarajan, Jure Leskovec, Christopher Potts, and Noah D Goodman. Causal interpretation of self-attention in pre-trained transformers. In Advances in Neural Information Processing Systems 36 (NeurIPS 2023), 2023. URL https://proceedings.neurips.cc/paper_files/paper/ 2023/file/642a321fba8a0f03765318e629cb93ea-Paper-Conference.pdf.A paper with this title exists at the given URL, but the authors don't match.MANTA: Cross-Modal Semantic Alignment and Information-Theoretic Optimization for Long-form Multimodal Understanding3.0MANTA: Cross-Modal Semantic Alignment and Information-Theoretic Optimization for Long-form Multimodal Understanding | OpenReviewhttps://app.gptzero.me/documents/381ed9a6-b168-4cd0-81ad-1f50139c0737/shareGuy Dove. Language as a cognitive tool to imagine goals in curiosity-driven exploration. Nature Communications, 13(1):1-14, 2022.An article with this title exists, but author and publication don't match.LOSI: Improving Multi-agent Reinforcement Learning via Latent Opponent Strategy Identification3.0LOSI: Improving Multi-agent Reinforcement Learning via Latent Opponent Strategy Identification | OpenReviewhttps://app.gptzero.me/documents/53e86e4b-a7e2-48d0-976b-240bfc412836/shareJing Liang, Fan Zhou, Shuying Li, Jun Chen, Guandong Zhou, Huaiming Xu, and Xin Li. Learning opponent behavior for robust cooperation in multi-agent reinforcement learning. IEEE Transactions on Cybernetics, 53(12):7527-7540, 2023.No MatchThe Dynamic Interaction Field Transformer: A Universal, Tokenizer-Free Language Architecture3.0The Dynamic Interaction Field Transformer: A Universal, Tokenizer-Free Language Architecture | OpenReviewhttps://app.gptzero.me/documents/80fd90a6-c99e-4c31-af72-0da9e90949f6/shareKaj Bostrom and Greg Durrett. Byte-level representation learning for multi-lingual named entity recognition. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 4617-4627, 2020.No MatchStrategema: Probabilistic Analysis of Adversarial Multi-Agent Behavior with LLMs in Social Deduction Games3.0Strategema: Probabilistic Analysis of Adversarial Multi-Agent Behavior with LLMs in Social Deduction Games | OpenReviewhttps://app.gptzero.me/documents/1155e8a8-f679-4942-8fd9-c47fb64ad967/shareTom Eccles, Jeffrey Tweedale, and Yvette Izza. Let's pretend: A study of negotiation with autonomous agents. In 2009 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology (WI-IAT), volume 3, pp. 449-452. IEEE, 2009.No MatchUnderstanding Transformer Architecture through Continuous Dynamics: A Partial Differential Equation Perspective3.0Understanding Transformer Architecture through Continuous Dynamics: A Partial Differential Equation Perspective | OpenReviewhttps://app.gptzero.me/documents/460a1a23-1a97-482a-9759-ade855a4a0b4/shareZijie J Wang, Yuhao Choi, and Dongyeop Wei. On the identity of the representation learned by pre-trained language models. arXiv preprint arXiv:2109.01819, 2021.No MatchDiffusion Aligned Embeddings2.8Diffusion Aligned Embeddings | OpenReviewhttps://app.gptzero.me/documents/3d95a003-06c6-4233-881b-03b1e29b4ba2/shareYujia Wang, Hu Huang, Cynthia Rudin, and Yaron Shaposhnik. Pacmap: Dimension reduction using pairwise controlled manifold approximation projection. Machine Learning, 110:559-590, 2021.A similar paper with two matching authors exists, but the other authors, title, and journal are wrong.Leveraging NLLB for Low-Resource Bidirectional Amharic – Afan Oromo Machine Translation2.5Leveraging NLLB for Low-Resource Bidirectional Amharic – Afan Oromo Machine Translation | Open Reviewhttps://app.gptzero.me/documents/813da6e2-f7e8-4c95-bdd8-7d29b8e4b641/shareAtnafa L. Tonja, Gebremedhin Gebremeskel, and Seid M. Yimam. Evaluating machine translation systems for ethiopian languages: A case study of amharic and afan oromo. Journal of Natural Language Engineering, 29(3):456-478, 2023.No MatchCertified Robustness Training: Closed-Form Certificates via CROWN2.5Certified Robustness Training: Closed-Form Certificates via CROWN | OpenReviewhttps://app.gptzero.me/documents/53b60ef5-2ebf-403e-8123-3a9bb2da0f33/shareHuan Zhang, Hongge Chen, Chaowei Xiao, and Bo Zhang. Towards deeper and better certified defenses against adversarial attacks. In International Conference on Learning Representations, 2019. URL https://openreview.net/forum?id=rJgG92A2mNo MatchContext-Aware Input Switching in Mobile Devices: A Multi-Language, Emoji-Integrated Typing System2.5Context-Aware Input Switching in Mobile Devices: A Multi-Language, Emoji-Integrated Typing System | OpenReviewhttps://app.gptzero.me/documents/68998766-49c3-4269-9eca-3b6a76ed68b4/shareIshan Tarunesh, Syama Sundar Picked, Sai Krishna Bhat, and Monojit Choudhury. Machine translation for code-switching: A systematic literature review. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics, pp. 3654-3670, 2021.Partial match to this article, but authors, title, and metadata is largely wrong.Five-Mode Tucker-LoRA for Video Diffusion on Conv3D Backbones2.5Five-Mode Tucker-LoRA for Video Diffusion on Conv3D Backbones | OpenReviewhttps://app.gptzero.me/documents/eb0fd660-ed00-4769-a940-3d093d4f1ec1/shareShengming Chen, Yuxin Wang, et al. Videocrafter: Open diffusion models for high-quality video generation. arXiv preprint arXiv:2305.07932, 2023b.A paper with the same title exists, but the authors and arXiv ID are wrong.Activation-Guided Regularization: Improving Deep Classifiers using Feature-Space Regularization with Dynamic Prototypes2.5Activation-Guided Regularization: Improving Deep Classifiers using Feature-Space Regularization with Dynamic Prototypes | OpenReviewhttps://app.gptzero.me/documents/4031111e-24ef-4e06-908e-18ab99b08932/shareWentao Cheng and Tong Zhang. Improving deep learning for classification with unknown label noise. In International Conference on Machine Learning, pp. 6059-6081. PMLR, 2023.A similar paper exists.Sparse-Smooth Decomposition for Nonlinear Industrial Time Series Forecasting2.5Sparse-Smooth Decomposition for Nonlinear Industrial Time Series Forecasting | OpenReviewhttps://app.gptzero.me/documents/c01ad49e-a788-4916-a6ee-f43314d14676/shareYutian Chen, Kun Zhang, Jonas Peters, and Bernhard Schölkopf. Causal discovery and inference for nonstationary systems. Journal of Machine Learning Research, 22(103):1-72, 2021.No MatchPDE-Transformer: A Continuous Dynamical Systems Approach to Sequence Modeling2.0PDE-Transformer: A Continuous Dynamical Systems Approach to Sequence Modeling | OpenReviewhttps://app.gptzero.me/documents/ba257eea-e86c-4276-84c0-08b7465e1e3e/share
Xuechen Li, Juntang Zhuang, Yifan Ding, Zhaozong Jin, Yun chen Chen, and Stefanie Jegelka. Scalable gradients for stochastic differential equations. In Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics (AISTATS 2020), volume 108 of Proceedings of Machine Learning Research, pp. 3898-3908, 2020.The paper exists and the first author is correct but all other authors and the page range are wrongSAFE-LLM: A Unified Framework for Reliable, Safe, And Secure Evaluation of Large Language Models2.0SAFE-LLM: A Unified Framework for Reliable, Safe, And Secure Evaluation of Large Language Models | OpenReviewhttps://app.gptzero.me/documents/05ee7ff4-40e2-48b7-b5bd-8c307d7db669/shareKuhn, J., et al. Semantic Entropy for Hallucination Detection. ACL 2023.A similar paper with different authors can be found here.PIPA: An Agent for Protein Interaction Identification and Perturbation Analysis2.0PIPA: An Agent for Protein Interaction Identification and Perturbation Analysis | OpenReviewhttps://app.gptzero.me/documents/5031a806-1271-4fd3-b333-2554f47cb9fa/shareAlex Brown et al. Autonomous scientific experimentation at the advanced light source using language-model-driven agents. Nature Communications, 16:7001, 2025.No MatchTyped Chain-of-Thought: A Curry-Howard Framework for Verifying LLM Reasoning2.0Typed Chain-of-Thought: A Curry-Howard Framework for Verifying LLM Reasoning | OpenReviewhttps://app.gptzero.me/documents/9d2e3239-99db-4712-be7f-e032156d92a5/shareDeepMind. Gemma scope: Scaling mechanistic interpretability to chain of thought. DeepMind Safety Blog, 2025. URL https://deepmindsafetyresearch.medium.com/ evaluating-and-monitoring-for-ai-scheming-8a7f2ce087f9. Discusses scaling mechanistic interpretability techniques to chain-of-thought and applications such as hallucination detection.ThA similar URL exists, and the title is similar to this blog. However, no exact match exists.Graph-Based Operator Learning from Limited Data on Irregular Domains2.0Graph-Based Operator Learning from Limited Data on Irregular Domains | OpenReviewhttps://app.gptzero.me/documents/6c52217f-fb88-4bd8-85aa-bd546e1fa88c/shareLiu, Y., Lütjens, B., Azizzadenesheli, K., and Anandkumar, A. (2022). U-netformer: A u-net style transformer for solving pdes. arXiv preprint arXiv:2206.11832.No MatchKARMA: Knowledge-Aware Reward Mechanism Adjustment via Causal AI2.0KARMA: Knowledge-Aware Reward Mechanism Adjustment via Causal AI | OpenReviewhttps://app.gptzero.me/documents/92b6492c-68ad-41a3-ae35-628d67f053e0/shareReinaldo A. C. Bianchi, Luis A. Celiberto Jr, and Ramon Lopez de Mantaras. Knowledge-based reinforcement learning: A survey. Journal of Artificial Intelligence Research, 62:215-261, 2018.No MatchMicroarchitecture Is Destiny: Performance and Accuracy of Quantized LLMs on Consumer Hardware2.0Microarchitecture Is Destiny: Performance and Accuracy of Quantized LLMs on Consumer Hardware | OpenReviewhttps://app.gptzero.me/documents/4504a39a-af72-41ab-9679-6f6a017a3275/shareZhihang Jiang, Dingkang Wang, Yao Li, et al. Fp6-llm: Efficient llm serving through fp6-centric co-design. arXiv preprint arXiv:2401.14112, 2024.the arXiv ID corresponds with a very similar paper, but the authors are wrong and the title is altered.Decoupling of Experts: A Knowledge-Driven Architecture for Efficient LLMs1.6Decoupling of Experts: A Knowledge-Driven Architecture for Efficient LLMs | OpenReviewhttps://app.gptzero.me/documents/74eade70-da36-4635-8749-5e1d04748b6d/shareH Zhang, Y L, X W, Y Z, X Z, H W, X H, K G, Z W, H W, H C, H L, and J W. Matrix data pile: A trillion-tokenscale datasets for llm pre-training. arXiv preprint arXiv:2408.12151, 2024.No Match; arxiv is is unrelatedQUART: Agentic Reasoning To Discover Missing Knowledge in Multi-Domain Temporal Data.1.5QUART: Agentic Reasoning To Discover Missing Knowledge in Multi-Domain Temporal Data. | OpenReviewhttps://app.gptzero.me/documents/c6f30343-3948-4c07-b7de-6b1407d5daa6/shareMeera Jain and Albert Chen. Explainable ai techniques for medical applications: A comprehensive review. AI in Healthcare, 5:22-37, 2024.No MatchFrom Physics-Informed Models to Deep Learning: Reproducible AI Frameworks for Climate Resilience and Policy Alignment1.5From Physics-Informed Models to Deep Learning: Reproducible AI Frameworks for Climate Resilience and Policy Alignment | OpenReviewhttps://app.gptzero.me/documents/a7ed6c42-4349-4b45-a356-0e325090e5af/shareMIT Climate Group. A cautionary tale for deep learning in climate science. https://example. com, 2019.The title matches this paper, but the citation is obviously hallucinated.A superpersuasive autonomous policy debating system1.5A superpersuasive autonomous policy debating system | OpenReviewhttps://app.gptzero.me/documents/b792a4de-baa8-47d4-b880-87b330a482ce/shareRoy Bar-Haim, Shachar Bhattacharya, Michal Jacovi, Yosi Mass, Matan Orbach, Eyal Sliwowicz, and Noam Slonim. Key point analysis via contrastive learning and extractive argument summarization. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 7953-7962, Online and Punta Cana, Dominican Republic, November 2021a. Association for Computational Linguistics. doi: 10.18653/v1/2021.emnlp-main.629. URL https://aclanthology.org/2021.emnlp-main. 629.A paper with the same title exists, but the authors and URL are wrong.AnveshanaAI: A Multimodal Platform for Adaptive AI/ML Education Through Automated Question Generation and Interactive Assessment1.5AnveshanaAI: A Multimodal Platform for Adaptive AI/ML Education Through Automated Question Generation and Interactive Assessment | OpenReviewhttps://app.gptzero.me/documents/720d6d24-2223-4e0e-95b9-6dfce674f8c7/shareShiyang Liu, Hongyi Xu, and Min Chen. Measuring and reducing perplexity in large-scale llms. arXiv preprint arXiv:2309.12345, 2023.No MatchAI-Assisted Medical Triage Assistant1.0AI-Assisted Medical Triage Assistant | OpenReviewhttps://app.gptzero.me/documents/391b5d76-929a-4f3f-addf-31f6993726f2/share[3] K. Arnold, J. Smith, and A. Doe. Variability in triage decision making. Resuscitation, 85:12341239, 2014.No MatchDeciphering Cross-Modal Feature Interactions in Multimodal AIGC Models: A Mechanistic Interpretability Approach0.67Deciphering Cross-Modal Feature Interactions in Multimodal AIGC Models: A Mechanistic Interpretability Approach | OpenReviewhttps://app.gptzero.me/documents/d4102812-01c4-45b2-aea8-59e467d31fd4/shareShuyang Basu, Sachin Y Gadre, Ameet Talwalkar, and Zico Kolter. Understanding multimodal llms: the mechanistic interpretability of llava in visual question answering. arXiv preprint arXiv:2411.17346, 2024.A paper with this title exists, but the authors and arXiv ID are wrong.Scalable Generative Modeling of Protein Ligand Trajectories via Graph Neural Diffusion Networks0.5Scalable Generative Modeling of Protein Ligand Trajectories via Graph Neural Diffusion Networks | OpenReviewhttps://app.gptzero.me/documents/32d43311-6e69-4b88-be99-682e4eb0c2cc/shareE. Brini, G. Jayachandran, and M. Karplus. Coarse-graining biomolecular simulations via statistical learning. J. Chem. Phys., 154:040901, 2021. There is no match for the title and authors, but the journal, volume, and year match this article
联系我们 contact @ memedata.com