cs.CL（2026-03-13）

📊 共 14 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) (10 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (3) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (10 篇)

#题目一句话要点标签🔗⭐
1 Continual Learning in Large Language Models: Methods, Challenges, and Opportunities 针对LLM的持续学习综述：方法、挑战与机遇 large language model
2 Neuron-Aware Data Selection In Instruction Tuning For Large Language Models 提出NAIT框架，通过神经元激活模式相似性进行指令调优数据选择，提升大语言模型性能。 large language model
3 DS$^2$-Instruct: Domain-Specific Data Synthesis for Large Language Models Instruction Tuning DS$^2$-Instruct：面向领域特定LLM指令调优的数据合成框架 large language model
4 SectEval: Evaluating the Latent Sectarian Preferences of Large Language Models SectEval：评估大型语言模型中潜在的宗派偏见 large language model ✅
5 ESG-Bench: Benchmarking Long-Context ESG Reports for Hallucination Mitigation 提出ESG-Bench基准数据集，用于评估和缓解大语言模型在ESG报告分析中的幻觉问题 large language model chain-of-thought
6 Adaptive Vision-Language Model Routing for Computer Use Agents 提出自适应VLM路由框架，优化计算机使用Agent的GUI操作成本与精度。 multimodal ✅
7 Experimental evidence of progressive ChatGPT models self-convergence 实验证据表明，ChatGPT模型存在因递归训练导致的自收敛现象 large language model
8 MetaKE: Meta-learning Aligned Knowledge Editing via Bi-level Optimization MetaKE：通过双层优化实现对齐知识编辑的元学习框架 large language model
9 From Text to Forecasts: Bridging Modality Gap with Temporal Evolution Semantic Space 提出TESS模型，通过时序演化语义空间弥合文本与时间序列预测的模态鸿沟 multimodal
10 CLARIN-PT-LDB: An Open LLM Leaderboard for Portuguese to assess Language, Culture and Civility 构建葡萄牙语开放LLM排行榜，评估语言、文化和文明程度 large language model ✅

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
11 Mending the Holes: Mitigating Reward Hacking in Reinforcement Learning for Multilingual Translation 提出WALAR方法，通过强化学习提升低资源多语翻译LLM性能。 reinforcement learning large language model
12 EvolveCoder: Evolving Test Cases via Adversarial Verification for Code Reinforcement Learning EvolveCoder：通过对抗验证进化测试用例，提升代码强化学习效果 reinforcement learning large language model
13 Rethinking Multiple-Choice Questions for RLVR: Unlocking Potential via Distractor Design 提出迭代干扰项构建(IDC)框架，提升RLVR中多选题的推理能力。 reinforcement learning large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
14 Expert Pyramid Tuning: Efficient Parameter Fine-Tuning for Expertise-Driven Task Allocation 提出专家金字塔调优(EPT)，通过多尺度特征金字塔提升参数高效微调性能。 manipulation

⬅️ 返回 cs.CL 首页 · 🏠 返回主页