cs.CL（2024-07-08）

📊 共 27 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) (25 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (2)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (25 篇)

#题目一句话要点标签🔗⭐
1 ANOLE: An Open, Autoregressive, Native Large Multimodal Models for Interleaved Image-Text Generation Anole：开放、自回归、原生的大型多模态模型，用于交错的图像-文本生成 large language model multimodal
2 An Empirical Study of Gendered Stereotypes in Emotional Attributes for Bangla in Multilingual Large Language Models 首次探究多语言大模型中孟加拉语情感属性的性别刻板印象 large language model
3 DebUnc: Improving Large Language Model Agent Communication With Uncertainty Metrics DebUnc：利用不确定性指标改进大语言模型Agent的通信 large language model ✅
4 Large Language Model Recall Uncertainty is Modulated by the Fan Effect 研究表明：大型语言模型表现出与人类相似的认知扇形效应，影响其回忆不确定性。 large language model
5 What's Wrong with Your Code Generated by Large Language Models? An Extensive Study 深入分析大型语言模型代码生成缺陷，并提出自纠错迭代方法 large language model
6 Merge, Ensemble, and Cooperate! A Survey on Collaborative Strategies in the Era of Large Language Models 综述大型语言模型协同策略：融合、集成与合作 large language model
7 Perceptions to Beliefs: Exploring Precursory Inferences for Theory of Mind in Large Language Models 提出PercepToM方法，提升大语言模型在心理理论任务中的表现 large language model
8 Limits to Predicting Online Speech Using Large Language Models 研究表明大型语言模型预测在线用户发言仍面临挑战，个性化建模至关重要 large language model
9 Large Language Models for Judicial Entity Extraction: A Comparative Study 利用大型语言模型进行司法实体抽取，提升法律文本信息处理效率。 large language model
10 Large Language Models Understand Layout 研究表明大语言模型具备理解空间布局的能力，并可用于提升视觉问答系统性能。 large language model
11 Do Multilingual Large Language Models Mitigate Stereotype Bias? 多语言训练有效缓解大型语言模型中的刻板印象偏见 large language model
12 Pruning Large Language Models to Intra-module Low-rank Architecture with Transitional Activations TransAct：通过模块内低秩架构剪枝LLM，降低激活值冗余 large language model
13 From Loops to Oops: Fallback Behaviors of Language Models Under Uncertainty 研究LLM不确定性下的回退行为，揭示模型能力与回退模式的关联。 large language model instruction following
14 LLaMAX: Scaling Linguistic Horizons of LLM by Enhancing Translation Capabilities Beyond 100 Languages LLaMAX：通过增强百余种语言的翻译能力，扩展LLM的语言边界 large language model foundation model ✅
15 When is the consistent prediction likely to be a correct prediction? 挑战自洽性理论：更长推理链而非最高频答案更可能正确 large language model chain-of-thought
16 When in Doubt, Cascade: Towards Building Efficient and Capable Guardrails 提出级联式Guardrail模型构建方法，提升效率与能力，用于检测LLM的不良输出。 large language model
17 CodeUpdateArena: Benchmarking Knowledge Editing on API Updates CodeUpdateArena：API更新场景下代码大模型知识编辑的基准测试 large language model
18 Using Grammar Masking to Ensure Syntactic Validity in LLM-based Modeling Tasks 提出语法掩码方法，确保LLM在建模任务中生成符合语法的模型 large language model
19 PAS: Data-Efficient Plug-and-Play Prompt Augmentation System PAS：数据高效的即插即用提示增强系统，提升LLM的易用性和有效性 large language model
20 KG-FPQ: Evaluating Factuality Hallucination in LLMs with Knowledge Graph-based False Premise Questions 提出KG-FPQ，利用知识图谱自动生成虚假前提问题，评估LLM的事实性幻觉 large language model ✅
21 Is GPT-4 Alone Sufficient for Automated Essay Scoring?: A Comparative Judgment Approach Based on Rater Cognition 提出基于比较判断的GPT-4自动作文评分方法，提升评分准确性 large language model
22 PsycoLLM: Enhancing LLM for Psychological Understanding and Evaluation PsycoLLM：增强LLM的心理理解与评估能力 large language model
23 InverseCoder: Self-improving Instruction-Tuned Code LLMs with Inverse-Instruct InverseCoder：利用逆向指令自提升指令调优的代码大语言模型 large language model
24 Open-world Multi-label Text Classification with Extremely Weak Supervision 提出X-MLClass，解决极弱监督下的开放世界多标签文本分类问题。 large language model
25 Generative Debunking of Climate Misinformation 提出一种基于大语言模型的框架，自动生成符合“真理三明治”结构的 climate change 错误信息辟谣内容。 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🔗⭐
26 Distilling System 2 into System 1 通过自监督蒸馏将System 2推理能力迁移至System 1，提升LLM效率 distillation large language model chain-of-thought
27 Retrieved In-Context Principles from Previous Mistakes 提出检索式上下文原则(RICP)，利用历史错误提升大语言模型推理能力 teacher-student large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页