cs.CL（2025-12-26）

📊 共 13 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (12) 支柱二：RL算法与架构 (RL & Architecture) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (12 篇)

#	题目	一句话要点	标签	🔗	⭐
1	Measuring Stability Beyond Accuracy in Small Open-Source Medical Large Language Models for Pediatric Endocrinology	评估儿科内分泌领域小型开源医学LLM的稳定性，超越传统准确率指标	large language model
2	Cross-Platform Evaluation of Large Language Model Safety in Pediatric Consultations: Evolution of Adversarial Robustness and the Scale Paradox	评估大语言模型在儿科咨询中的安全性，揭示对抗鲁棒性演变与规模悖论	large language model
3	Bounded Hyperbolic Tangent: A Stable and Efficient Alternative to Pre-Layer Normalization in Large Language Models	提出有界双曲正切(BHyT)，提升大语言模型训练稳定性和效率，替代Pre-LN。	large language model
4	TimeBill: Time-Budgeted Inference for Large Language Models	TimeBill：面向大语言模型的时间预算推理框架，提升任务完成率和响应性能。	large language model
5	Knowledge Reasoning of Large Language Models Integrating Graph-Structured Information for Pest and Disease Control in Tobacco	提出融合图结构信息的大语言模型，用于烟草病虫害防治的知识推理	large language model
6	Towards Efficient Post-Training via Fourier-Driven Adapter Architectures	提出基于傅里叶变换的Adapter架构FAA，用于高效微调大型预训练语言模型。	large language model
7	CricBench: A Multilingual Benchmark for Evaluating LLMs in Cricket Analytics	CricBench：一个用于评估LLM在板球分析中性能的多语言基准测试	large language model
8	Bridging the Copyright Gap: Do Large Vision-Language Models Recognize and Respect Copyrighted Content?	评估大型视觉语言模型版权意识，提出工具增强防御框架以降低侵权风险	multimodal
9	Context as a Tool: Context Management for Long-Horizon SWE-Agents	提出CAT框架，通过可调用工具管理上下文，提升长程软件工程Agent性能。	large language model
10	Broken Words, Broken Performance: Effect of Tokenization on Performance of LLMs	研究表明：LLM分词方式影响性能，提出惩罚函数量化分词质量	large language model
11	Method Decoration (DeMe): A Framework for LLM-Driven Adaptive Method Generation in Dynamic IoT Environments	提出DeMe框架，利用LLM驱动IoT环境下的自适应方法生成	large language model
12	On The Conceptualization and Societal Impact of Cross-Cultural Bias	分析跨文化偏见文献，倡导语言技术社会影响评估	large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
13	StatLLaMA: A multi-stage training framework for building a domain-optimized statistical language model	StatLLaMA：一种用于构建领域优化统计语言模型的多阶段训练框架	reinforcement learning RLHF direct preference optimization	✅

⬅️ 返回 cs.CL 首页 · 🏠 返回主页