cs.CL（2024-11-25）

📊 共 31 篇论文

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (21) 支柱二：RL算法与架构 (RL & Architecture) (8) 支柱一：机器人控制 (Robot Control) (2)

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

#题目一句话要点标签🔗⭐
1 BayLing 2: A Multilingual Large Language Model with Efficient Language Alignment BayLing 2：通过高效语言对齐增强多语言大语言模型能力 large language model foundation model
2 Do Large Language Models Perform Latent Multi-Hop Reasoning without Exploiting Shortcuts? 提出SOCRATES数据集，评估大语言模型在无捷径条件下潜在多跳推理能力。 large language model chain-of-thought
3 AtomR: Atomic Operator-Empowered Large Language Models for Heterogeneous Knowledge Reasoning AtomR：原子操作赋能大语言模型，用于异构知识推理 large language model chain-of-thought
4 TransCompressor: LLM-Powered Multimodal Data Compression for Smart Transportation TransCompressor：利用LLM进行智能交通多模态数据压缩与重建 large language model multimodal
5 Can AI grade your essays? A comparative analysis of large language models and teacher ratings in multidimensional essay scoring 评估LLM在多维度作文评分中的表现，探索AI辅助教师的新途径 large language model
6 Enhancing Answer Reliability Through Inter-Model Consensus of Large Language Models 提出基于多大型语言模型共识的框架，提升复杂问题回答的可靠性。 large language model
7 EnStack: An Ensemble Stacking Framework of Large Language Models for Enhanced Vulnerability Detection in Source Code EnStack：一种基于大语言模型集成堆叠的源代码漏洞检测框架 large language model
8 DoubleCCA: Improving Foundation Model Group Robustness with Random Sentence Embeddings 提出DoubleCCA方法，利用随机句子嵌入增强基础模型对群体偏见的鲁棒性。 foundation model
9 Lessons from Studying Two-Hop Latent Reasoning 研究表明大语言模型具备潜在的双跳推理能力，但事实组合仍具挑战 large language model chain-of-thought
10 Enhancing LLM Reasoning via Critique Models with Test-Time and Training-Time Supervision 提出基于批判模型的LLM推理增强方法，提升复杂推理任务性能 large language model
11 Teaching Smaller Language Models To Generalise To Unseen Compositional Questions (Full Thesis) 提出检索增强训练数据集(RATD)和知识融合方法，提升小模型在复杂推理问答上的泛化能力。 large language model
12 What can LLM tell us about cities? 利用大型语言模型探索城市知识：一种数据驱动的城市研究新范式 large language model
13 Parameter Efficient Instruction Tuning: An Empirical Study 参数高效指令调优实证研究：揭示LoRA和Adapter的性能边界与适用场景 instruction following
14 LLM Augmentations to support Analytical Reasoning over Multiple Documents 提出动态证据树以增强多文档分析推理能力 large language model
15 Profiling Bias in LLMs: Stereotype Dimensions in Contextual Word Embeddings 提出基于刻板印象维度偏见剖析方法，用于评估大型语言模型中的性别偏见。 large language model
16 Learning by Analogy: Enhancing Few-Shot Prompting for Math Word Problem Solving with Computational Graph-Based Retrieval 提出基于计算图检索的类比学习方法，提升LLM在数学应用题上的少样本提示能力 large language model
17 FineWeb-zhtw: Scalable Curation of Traditional Chinese Text Data from the Web FineWeb-zhtw：构建大规模高质量繁体中文网络文本数据集 large language model
18 Multi-modal Retrieval Augmented Multi-modal Generation: Datasets, Evaluation Metrics and Strong Baselines 提出多模态检索增强多模态生成框架M²RAG，并构建数据集、评估指标和基线模型。 foundation model
19 NormXLogit: The Head-on-Top Never Lies 提出NormXLogit，一种模型无关的LLM可解释性方法，提升token重要性评估的忠实性。 large language model
20 MH-MoE: Multi-Head Mixture-of-Experts 提出MH-MoE，利用多头机制提升稀疏MoE模型的性能，同时保持参数量和计算量不变。 large language model
21 SAGEval: The frontiers of Satisfactory Agent based NLG Evaluation for reference-free open-ended text 提出SAGEval框架，利用批判Agent提升无参考开放文本生成评估质量。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
22 Dynamic Self-Distillation via Previous Mini-batches for Fine-tuning Small Language Models 提出DynSDPB，通过前一批次动态自蒸馏微调小型语言模型，无需复杂教师模型。 distillation large language model
23 Self-Generated Critiques Boost Reward Modeling for Language Models Critic-RM：利用自生成评判提升语言模型奖励建模能力 reinforcement learning RLHF large language model
24 Preference Optimization for Reasoning with Pseudo Feedback 提出基于伪反馈的偏好优化方法，提升LLM在数学推理和代码生成任务上的性能。 DPO direct preference optimization large language model
25 KL-geodesics flow matching with a novel sampling scheme 提出条件流匹配方法以提升文本生成性能 flow matching
26 O1 Replication Journey -- Part 2: Surpassing O1-preview through Simple Distillation, Big Progress or Bitter Lesson? 通过简单蒸馏超越O1-preview：数学推理与泛化能力提升 distillation
27 When Babies Teach Babies: Can student knowledge sharing outperform Teacher-Guided Distillation on small datasets? 提出一种无教师的加权互学习方法，在小数据集上实现高效语言模型预训练。 distillation
28 SHuBERT: Self-Supervised Sign Language Representation Learning via Multi-Stream Cluster Prediction SHuBERT：通过多流聚类预测实现手语自监督表征学习 representation learning
29 Contrastive Multi-graph Learning with Neighbor Hierarchical Sifting for Semi-supervised Text Classification 提出ConNHS，通过对比多图学习和邻域分层筛选改进半监督文本分类 representation learning contrastive learning

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
30 Preventing Jailbreak Prompts as Malicious Tools for Cybercriminals: A Cyber Defense Perspective 从网络防御角度分析并预防利用Jailbreak Prompt的网络犯罪 manipulation large language model
31 Transparent Neighborhood Approximation for Text Classifier Explanation 提出XPROB，一种基于概率编辑的透明文本分类器解释方法 manipulation

⬅️ 返回 cs.CL 首页 · 🏠 返回主页