cs.CL（2026-02-20）

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

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

#	题目	一句话要点	标签	🔗	⭐
1	FENCE: A Financial and Multimodal Jailbreak Detection Dataset	提出FENCE：金融多模态越狱检测数据集，提升金融领域AI系统安全性	large language model multimodal
2	PolyFrame at MWE-2026 AdMIRe 2: When Words Are Not Enough: Multimodal Idiom Disambiguation	PolyFrame通过轻量级模块提升多模态成语歧义消解性能	multimodal zero-shot transfer
3	Click it or Leave it: Detecting and Spoiling Clickbait with Informativeness Measures and Large Language Models	提出一种融合信息量化特征与大语言模型的混合方法，用于检测并消除网络点击诱饵。	large language model
4	SPQ: An Ensemble Technique for Large Language Model Compression	SPQ：一种用于大语言模型压缩的集成技术，在内存受限环境下实现高效部署。	large language model	✅
5	Analyzing LLM Instruction Optimization for Tabular Fact Verification	针对表格事实核查，提出基于DSPy框架的大语言模型指令优化方法	large language model chain-of-thought
6	VeriSoftBench: Repository-Scale Formal Verification Benchmarks for Lean	VeriSoftBench：面向Lean的形式化验证，仓库级规模的基准测试集	large language model	✅
7	Simplifying Outcomes of Language Model Component Analyses with ELIA	ELIA：利用AI解释简化语言模型内部机制分析，提升可访问性	large language model
8	Agentic Adversarial QA for Improving Domain-Specific LLMs	提出Agentic对抗问答框架，提升领域特定LLM的性能和样本效率。	large language model
9	Vichara: Appellate Judgment Prediction and Explanation for the Indian Judicial System	Vichara：针对印度司法系统的上诉判决预测与解释框架	large language model
10	Thinking by Subtraction: Confidence-Driven Contrastive Decoding for LLM Reasoning	提出基于置信度对比解码的LLM推理方法，提升推理可靠性并减少输出冗余。	large language model	✅
11	The Statistical Signature of LLMs	利用无损压缩识别LLM文本的统计特征，揭示生成模型在不同信息生态中的结构规律。	large language model
12	Detecting Contextual Hallucinations in LLMs with Frequency-Aware Attention	提出基于频率感知的注意力机制，用于检测LLM在上下文生成中的幻觉问题。	large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
13	VIRAASAT: Traversing Novel Paths for Indian Cultural Reasoning	VIRAASAT：提出印度文化推理数据集与SCoM框架，提升文化知识推理能力	manipulation large language model chain-of-thought

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

#	题目	一句话要点	标签	🔗	⭐
14	DP-RFT: Learning to Generate Synthetic Text via Differentially Private Reinforcement Fine-Tuning	提出DP-RFT，通过差分隐私强化微调生成高质量合成文本数据	reinforcement learning PPO large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页