Loading...
###摘要药物发现是生物医学自然语言处理（NLP）中的一项关键任务，然而可解释的药物发现仍有待深入探索。与此同时，大语言模型（LLMs）在自然语言理解和生成方面展现出了卓越的能力。利用大语言模型进行可解释的药物发现，有望改进下游任务和
Re-evaluating Open-ended Evaluation of Large Language Models
###摘要传统的评估方式主要聚焦于针对特定技能对候选对象进行排名。而诸如大语言模型（LLMs）这样的现代通用模型，显然超越了这种范式。开放式评估系统应运而生，在该系统中，候选模型会依据用户提交的提示进行比较，成为了一种流行的解决方案。
HaLoRA: Hardware-aware Low-Rank Adaptation for Large Language Models Based on Hybrid Compute-in-Memo
###摘要低秩自适应（LoRA）是一种主流的参数高效微调方法，用于使大语言模型（LLMs）适应下游任务。在本文中，我们首次提出将LoRA微调后的大语言模型部署在混合内存计算（CIM）架构上（即，预训练权重存储在RRAM中，Lo
Secure Speculative Decoding for Large Language Models
#SecureSpeculativeDecodingforLargeLanguageModels论文总结与翻译>论文信息：arXiv:2610.08678，已被**IEEES&P2027**接收，18页；作者
Selective Critique for Cost‑Aware LLM Agents in Long‑Horizon Decision Making
#SelectiveCritiqueforCost‑AwareLLMAgentsinLong‑HorizonDecisionMaking>arXiv:2610.07335|NeurIPS2026##文章主要
ThinkFuse: Trajectory‑Aware Test‑Time Fusion for Small Reasoning Models
#ThinkFuse:Trajectory‑AwareTest‑TimeFusionforSmallReasoningModels>arXiv:2610.07803，接收于EMNLP2026Findings##
Learning to Retrieve via Reinforcement Learning in Embedding Space
LearningtoRetrieveviaReinforcementLearninginEmbeddingSpace论文总结与翻译论文arXiv地址：https://arxiv.org/abs/2610.07731
Foresight-over-Graph: Reasoning Beyond Local Horizons for Knowledge Base Question Answering
>说明：当前仅获取到论文摘要元数据，**无法拿到Introduction与Conclusion原文**，因此仅完成：文章主要内容&创新点总结、Abstract中英对照翻译；Introduction、Conclusion部分因缺少全文文本无
SquidAgent: Parallelize Wisely, Coordinate Efficiently
#SquidAgent:ParallelizeWisely,CoordinateEfficiently论文总结与翻译>论文信息：arXiv:2610.08647，NeurIPS2026接收，37页（含附录）##一、文
The Model Plants the Trigger: Answer‑Side Backdoor Attacks in Multi‑Turn Large Language Models
#TheModelPlantstheTrigger:Answer‑SideBackdoorAttacksinMulti‑TurnLargeLanguageModels>Paper:arXiv:2610.077
COMPASS: Finding Where Reasoning Lives in Language Models（arXiv:2610.07469）
#COMPASS:FindingWhereReasoningLivesinLanguageModels（arXiv:2610.07469）>论文链接：https://arxiv.org/abs/2610.07469#
WASD: Wasserstein‑based Knowledge Distillation for Large Language Models 论文总结与翻译
#WASD:Wasserstein‑basedKnowledgeDistillationforLargeLanguageModels论文总结与翻译>论文信息：arXiv:2610.07706，NeurIPS2026
Safeguarding LLMs via Model‑Agnostic Latent Safety Signals from Dark Knowledge 论文阅读笔记
#SafeguardingLLMsviaModel‑AgnosticLatentSafetySignalsfromDarkKnowledge论文阅读笔记>论文arXiv链接：[https://arxiv.org/a
Beyond Refusal Patterns 论文总结与翻译
#BeyondRefusalPatterns论文总结与翻译>论文信息：arXiv:2610.07023|BeyondRefusalPatterns:Safe‑RoleInternalizationforRobus
Distributionally Robust Mixture‑of‑Experts Training（DRMoET）论文总结与翻译
#DistributionallyRobustMixture‑of‑ExpertsTraining（DRMoET）论文总结与翻译>论文信息：arXiv:2610.07207，NeurIPS2026，作者：XinTeng,M
CoDe‑LoRA 论文阅读总结
#CoDe‑LoRA论文阅读总结>论文信息：CoDe‑LoRA:MitigatingtheOrthogonalityDilemmainContinualLearningofLLMsviaKnowledgeCo
Dynamic Positional Attention Modulation 论文阅读笔记
#DynamicPositionalAttentionModulation论文阅读笔记>论文信息：arXiv:2610.07848，KDD2026，标题：*DynamicPositionalAttentionModul
Illusory Pattern Perception Drives Spurious Inference in Large Language Models 论文总结与译文
#IllusoryPatternPerceptionDrivesSpuriousInferenceinLargeLanguageModels论文总结与译文>论文信息：arXiv:2610.07791，NeurIPS
Understanding and Enhancing Backdoor Persistency in LLM Agent Post‑Training 论文总结与翻译
#UnderstandingandEnhancingBackdoorPersistencyinLLMAgentPost‑Training论文总结与翻译>论文arXiv：https://arxiv.org/abs/26
MILO: Efficient Many‑shot In‑Context Learning with Block‑wise Low‑rank Compression
>说明：arXiv摘要页面仅公开**Abstract**，未直接展示Introduction、Conclusion完整原文；PDF需要读取完整论文。当前可获取完整的只有Abstract，因此我完成：论文整体总结、创新点、Abstract中
