cs.CL（2026-03-17）

📊 共 21 篇论文 | 🔗 3 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (17 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (4)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (17 篇)

#题目一句话要点标签🔗⭐
1 Probing Cultural Signals in Large Language Models through Author Profiling 通过作者画像探测大型语言模型中的文化信号 large language model ✅
2 Omanic: Towards Step-wise Evaluation of Multi-hop Reasoning in Large Language Models Omanic：用于大语言模型多跳推理分步评估的基准数据集 large language model ✅
3 Arabic Morphosyntactic Tagging and Dependency Parsing with Large Language Models 利用大型语言模型进行阿拉伯语词法句法标注和依存句法分析 large language model
4 DynHD: Hallucination Detection for Diffusion Large Language Models via Denoising Dynamics Deviation Learning DynHD：通过去噪动态偏差学习检测扩散大语言模型的幻觉 large language model
5 Structured Semantic Cloaking for Jailbreak Attacks on Large Language Models 提出结构化语义伪装S2C，用于绕过大型语言模型的越狱攻击防御。 large language model
6 BATQuant: Outlier-resilient MXFP4 Quantization via Learnable Block-wise Optimization BATQuant：提出一种基于可学习分块优化的、对异常值鲁棒的MXFP4量化方法 large language model multimodal
7 Pre-training LLM without Learning Rate Decay Enhances Supervised Fine-Tuning 预训练LLM不使用学习率衰减可增强监督微调性能 large language model
8 Who Benchmarks the Benchmarks? A Case Study of LLM Evaluation in Icelandic 针对冰岛语LLM评测，揭示合成/机器翻译数据偏差问题，呼吁改进低资源语言评测方法。 large language model
9 Frequency Matters: Fast Model-Agnostic Data Curation for Pruning and Quantization 提出ZipCal：一种快速、模型无关的数据筛选方法，用于模型剪枝和量化。 large language model
10 Chronos: Temporal-Aware Conversational Agents with Structured Event Retrieval for Long-Term Memory Chronos：利用结构化事件检索和时间感知能力，增强对话Agent的长期记忆。 large language model
11 Good Arguments Against the People Pleasers: How Reasoning Mitigates (Yet Masks) LLM Sycophancy 研究表明思维链推理虽能降低大语言模型谄媚，但也会掩盖其潜在倾向。 chain-of-thought
12 AdaMem: Adaptive User-Centric Memory for Long-Horizon Dialogue Agents AdaMem：面向长程对话Agent的自适应用户中心记忆框架 large language model
13 VQKV: High-Fidelity and High-Ratio Cache Compression via Vector-Quantization VQKV：通过向量量化实现高保真和高压缩比的KV缓存压缩 large language model
14 SpecSteer: Synergizing Local Context and Global Reasoning for Efficient Personalized Generation 提出SpecSteer以解决个性化生成中的隐私与推理能力问题 large language model
15 Parametric Social Identity Injection and Diversification in Public Opinion Simulation 提出参数化社会身份注入方法，提升LLM在公共舆论模拟中的多样性 large language model ✅
16 SIA: A Synthesize-Inject-Align Framework for Knowledge-Grounded and Secure E-commerce Search LLMs with Industrial Deployment 提出SIA框架，解决电商搜索LLM的知识幻觉和安全漏洞问题，已在京东部署。 large language model
17 ClaimFlow: Tracing the Evolution of Scientific Claims in NLP 提出ClaimFlow，追踪NLP领域科学主张的演变，并构建主张关系分类任务。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#题目一句话要点标签🔗⭐
18 EmoLLM: Appraisal-Grounded Cognitive-Emotional Co-Reasoning in Large Language Models EmoLLM：基于评估理论的大语言模型认知-情感协同推理框架 reinforcement learning large language model
19 ASDA: Automated Skill Distillation and Adaptation for Financial Reasoning ASDA：金融推理的自动化技能蒸馏与适配框架 distillation large language model
20 PlotTwist: A Creative Plot Generation Framework with Small Language Models PlotTwist：一种利用小型语言模型进行创造性情节生成的框架，性能媲美大型模型。 direct preference optimization large language model
21 Online Experiential Learning for Language Models 提出在线经验学习框架OEL，提升语言模型在真实部署环境中的性能 distillation large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页