cs.CL（2024-08-27）

📊 共 23 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (17 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (6)

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

#题目一句话要点标签🔗⭐
1 X-Reflect: Cross-Reflection Prompting for Multimodal Recommendation 提出X-Reflect，通过跨模态反思提示增强多模态推荐系统性能 large language model multimodal
2 Non-instructional Fine-tuning: Enabling Instruction-Following Capabilities in Pre-trained Language Models without Instruction-Following Data 提出非指令微调方法，无需指令数据即可提升预训练语言模型的指令遵循能力 large language model instruction following
3 Toward Large Language Models as a Therapeutic Tool: Comparing Prompting Techniques to Improve GPT-Delivered Problem-Solving Therapy 探索大型语言模型在问题解决疗法中的应用：提示工程提升GPT疗效 large language model
4 Advancing Adversarial Suffix Transfer Learning on Aligned Large Language Models 提出DeGCG框架，加速对齐大语言模型的对抗性后缀生成与迁移，提升安全性。 large language model
5 BaichuanSEED: Sharing the Potential of ExtensivE Data Collection and Deduplication by Introducing a Competitive Large Language Model Baseline BaichuanSEED：开源大规模数据处理流程，并验证其在7B LLM上的有效性 large language model
6 Large Language Models for Disease Diagnosis: A Scoping Review 综述：大型语言模型在疾病诊断中的应用与评估 large language model
7 A Survey of Large Language Models for European Languages 综述性研究：针对欧洲语言的大型语言模型及其构建增强方法 large language model
8 LyCon: Lyrics Reconstruction from the Bag-of-Words Using Large Language Models 提出LyCon：利用大型语言模型从词袋数据重建歌词，解决版权限制下的歌词研究难题 large language model
9 Zero-Shot Visual Reasoning by Vision-Language Models: Benchmarking and Analysis 通过合成数据集，系统性评测视觉语言模型在零样本视觉推理上的能力与局限性 large language model chain-of-thought
10 Implicit Geometry of Next-token Prediction: From Language Sparsity Patterns to Model Representations 通过分析Next-token预测的隐式几何结构，揭示语言稀疏模式与模型表示之间的关系 large language model
11 Can Unconfident LLM Annotations Be Used for Confident Conclusions? 提出置信度驱动推理，利用LLM标注与置信度指标优化人工标注，提升计算社会科学研究效率。 large language model
12 Nuance Matters: Probing Epistemic Consistency in Causal Reasoning 提出因果认知一致性评估框架，揭示LLM在细粒度因果推理中存在的认知不一致问题。 large language model
13 Awes, Laws, and Flaws From Today's LLM Research 分析2000+LLM研究，揭示伦理声明下降、LLM自评估上升等趋势，并提出改进建议。 large language model
14 AgentMonitor: A Plug-and-Play Framework for Predictive and Secure Multi-Agent Systems AgentMonitor：用于预测和安全多智能体系统的即插即用框架 large language model ✅
15 Writing in the Margins: Better Inference Pattern for Long Context Retrieval 提出WiM以优化长输入序列的检索任务 large language model ✅
16 AAVENUE: Detecting LLM Biases on NLU Tasks in AAVE via a Novel Benchmark AAVENUE：提出新基准评测LLM在AAVE的NLU任务上的偏差 large language model
17 PolicyLR: A Logic Representation For Privacy Policies 提出PolicyLR以解决隐私政策理解与分析问题 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
18 Inverse-Q*: Token Level Reinforcement Learning for Aligning Large Language Models Without Preference Data 提出Inverse-Q*，无需偏好数据即可进行token级强化学习对齐LLM reinforcement learning PPO RLHF
19 DualKanbaFormer: An Efficient Selective Sparse Framework for Multimodal Aspect-based Sentiment Analysis 提出DualKanbaFormer，用于高效选择性稀疏多模态情感分析。 Mamba state space model multimodal
20 Negation Blindness in Large Language Models: Unveiling the NO Syndrome in Image Generation 揭示大型语言模型在图像生成中存在的“否定盲区”问题（NO综合征） reinforcement learning large language model
21 Atoxia: Red-teaming Large Language Models with Target Toxic Answers Atoxia：利用目标有害答案进行大语言模型红队测试 reinforcement learning large language model
22 GSIFN: A Graph-Structured and Interlaced-Masked Multimodal Transformer-based Fusion Network for Multimodal Sentiment Analysis 提出GSIFN，利用图结构和交错掩码Transformer融合网络提升多模态情感分析性能。 representation learning multimodal
23 SpikingSSMs: Learning Long Sequences with Sparse and Parallel Spiking State Space Models 提出SpikingSSM，利用脉冲神经网络和状态空间模型进行高效长序列学习 SSM state space model large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页