cs.CL（2025-01-31）

📊 共 24 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) (21 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (2) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (21 篇)

#题目一句话要点标签🔗⭐
1 Large Language Models as Common-Sense Heuristics 利用大语言模型作为常识启发式，提升家庭环境下的规划任务成功率。 large language model
2 Benchmarking Gaslighting Negation Attacks Against Multimodal Large Language Models GaslightingBench：评估多模态大语言模型抵抗否定攻击的基准 large language model multimodal ✅
3 Do Large Multimodal Models Solve Caption Generation for Scientific Figures? Lessons Learned from SciCap Challenge 2023 SciCap Challenge 2023揭示大型多模态模型在科学图表标题生成中的能力与局限性 multimodal
4 mFollowIR: a Multilingual Benchmark for Instruction Following in Retrieval 提出mFollowIR多语言基准，评估检索模型在指令跟随方面的能力。 instruction following
5 Efficient Reasoning with Hidden Thinking 提出Heima框架，通过隐空间推理提升多模态大语言模型效率。 large language model multimodal chain-of-thought
6 Fairshare Data Pricing via Data Valuation for Large Language Models 提出Fairshare数据定价机制，解决LLM训练数据市场剥削性定价问题。 large language model
7 Resolving Editing-Unlearning Conflicts: A Knowledge Codebook Framework for Large Language Model Updating 提出LOKA：基于知识代码本框架解决大语言模型更新中的编辑-遗忘冲突 large language model
8 Large Language Models' Accuracy in Emulating Human Experts' Evaluation of Public Sentiments about Heated Tobacco Products on Social Media 利用大型语言模型模拟人类专家评估社交媒体上对加热烟草制品的情绪 large language model
9 Efficient Beam Search for Large Language Models Using Trie-Based Decoding 提出基于Trie树解码的高效Beam Search方法，解决大语言模型内存瓶颈。 large language model
10 Intrinsic Tensor Field Propagation in Large Language Models: A Novel Approach to Contextual Information Flow 提出内在张量场传播(ITFP)，增强语言模型长程依赖建模能力。 large language model
11 Text Data Augmentation for Large Language Models: A Comprehensive Survey of Methods, Challenges, and Opportunities 综述性研究：针对大语言模型的文本数据增强方法、挑战与机遇 large language model
12 Pheromone-based Learning of Optimal Reasoning Paths 提出基于蚁群优化的思维树方法(ACO-ToT)，提升LLM复杂推理能力 large language model chain-of-thought
13 Accelerating LLM Inference with Lossless Speculative Decoding Algorithms for Heterogeneous Vocabularies 提出无需同词表的无损推测解码算法，加速异构LLM推理。 large language model
14 Ensembles of Low-Rank Expert Adapters 提出ELREA框架，通过低秩专家适配器集成解决LLM微调中的任务冲突问题。 large language model
15 BTS: Harmonizing Specialized Experts into a Generalist LLM BTS：通过融合领域专家模型，高效构建通用大语言模型 large language model
16 Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming 提出宪法分类器防御通用越狱攻击，在数千小时对抗测试中表现出色 large language model
17 A Checks-and-Balances Framework for Context-Aware Ethical AI Alignment 提出基于三权分立的上下文感知伦理AI对齐框架，保障LLM的伦理行为。 large language model
18 Reward-Guided Speculative Decoding for Efficient LLM Reasoning 提出奖励引导的推测解码（RSD），提升大语言模型推理效率。 large language model ✅
19 Beyond checkmate: exploring the creative chokepoints in AI text 探索AI文本创作的薄弱环节：基于文本片段差异的人工智能文本检测 large language model ✅
20 Sparse Autoencoder Insights on Voice Embeddings 利用稀疏自编码器解析语音嵌入，揭示单义特征 large language model
21 LLM-based Affective Text Generation Quality Based on Different Quantization Values 研究量化精度对LLM情感文本生成质量的影响，探索资源效率与性能的权衡。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (2 篇)

#题目一句话要点标签🔗⭐
22 Structural Embedding Projection for Contextual Large Language Model Inference 提出结构化嵌入投影(SEP)以提升上下文大语言模型推理效率与连贯性。 representation learning large language model
23 DyPCL: Dynamic Phoneme-level Contrastive Learning for Dysarthric Speech Recognition 提出动态音素级对比学习(DyPCL)用于提升构音障碍语音识别性能 contrastive learning curriculum learning

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
24 Token Sampling Uncertainty Does Not Explain Homogeneity Bias in Large Language Models 研究表明Token采样不确定性无法解释大型语言模型中的同质性偏差 manipulation representation learning large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页