cs.CL（2025-11-21）

📊 共 25 篇论文 | 🔗 4 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (17 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (6 🔗1) 支柱六：视频提取与匹配 (Video Extraction) (1) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (17 篇)

#	题目	一句话要点	标签	🔗	⭐
1	Affective Multimodal Agents with Proactive Knowledge Grounding for Emotionally Aligned Marketing Dialogue	提出AffectMind，通过主动知识 grounding 实现情感对齐的多模态营销对话。	large language model multimodal
2	Don't Learn, Ground: A Case for Natural Language Inference with Visual Grounding	提出一种基于视觉 grounding 的零样本自然语言推理方法，提升模型鲁棒性。	multimodal visual grounding
3	Large Language Models for Sentiment Analysis to Detect Social Challenges: A Use Case with South African Languages	利用大型语言模型进行情感分析，以检测南非语言中的社会挑战。	large language model
4	R2Q: Towards Robust 2-Bit Large Language Models via Residual Refinement Quantization	提出R2Q：通过残差细化量化实现鲁棒的2比特大语言模型	large language model
5	Hallucinate Less by Thinking More: Aspect-Based Causal Abstention for Large Language Models	提出基于知识切面的因果消融框架ABCA，减少大语言模型的幻觉问题	large language model
6	Training Foundation Models on a Full-Stack AMD Platform: Compute, Networking, and System Design	在全栈AMD平台上训练基础模型，优化计算、网络和系统设计。	foundation model	✅
7	RoSA: Enhancing Parameter-Efficient Fine-Tuning via RoPE-aware Selective Adaptation in Large Language Models	RoSA：通过RoPE感知的选择性适配增强大语言模型的参数高效微调	large language model	✅
8	Supervised Fine Tuning of Large Language Models for Domain Specific Knowledge Graph Construction:A Case Study on Hunan's Historical Celebrities	针对领域知识图谱构建，提出基于监督微调的大语言模型方法	large language model
9	PoETa v2: Toward More Robust Evaluation of Large Language Models in Portuguese	PoETa v2：构建葡萄牙语LLM评测基准，促进更鲁棒的模型评估	large language model	✅
10	Asking LLMs to Verify First is Almost Free Lunch	提出Verification-First策略，以低成本提升LLM的推理能力	large language model chain-of-thought
11	A Multiscale Geometric Method for Capturing Relational Topic Alignment	提出多尺度几何方法以捕捉关系主题对齐问题	multimodal
12	PARROT: Persuasion and Agreement Robustness Rating of Output Truth -- A Sycophancy Robustness Benchmark for LLMs	PARROT：提出一种评估LLM在权威诱导下输出真值一致性的基准	large language model
13	The PLLuM Instruction Corpus	PLLuM项目发布指令数据集PLLuMIC，用于微调波兰语大型语言模型，并分析人工与合成指令的影响。	large language model
14	Identifying Quantum Structure in AI Language: Evidence for Evolutionary Convergence of Human and Artificial Cognition	大型语言模型概念组合测试揭示量子结构，或表明人类与AI认知的进化趋同	large language model
15	AutoLink: Autonomous Schema Exploration and Expansion for Scalable Schema Linking in Text-to-SQL at Scale	AutoLink：面向大规模Text-to-SQL的自主模式探索与扩展，实现可扩展的模式链接	large language model
16	MUCH: A Multilingual Claim Hallucination Benchmark	提出多语言声明幻觉基准MUCH，用于评估和提升LLM的声明级不确定性量化。	large language model
17	OmniScientist: Toward a Co-evolving Ecosystem of Human and AI Scientists	提出OmniScientist框架，构建人机协同的AI科学家共生生态系统	large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#	题目	一句话要点	标签	🔗	⭐
18	Learning to Compress: Unlocking the Potential of Large Language Models for Text Representation	提出LLM2Comp，利用上下文压缩预训练提升大语言模型文本表示能力	contrastive learning large language model	✅
19	E$^3$-Pruner: Towards Efficient, Economical, and Effective Layer Pruning for Large Language Models	E$^3$-Pruner：面向大语言模型的高效、经济、有效层剪枝框架	distillation large language model
20	Masked-and-Reordered Self-Supervision for Reinforcement Learning from Verifiable Rewards	提出MR-RLVR，通过掩码和重排序自监督提升RLVR在可验证奖励下的数学推理能力	reinforcement learning large language model
21	Decoding inner speech with an end-to-end brain-to-text neural interface	提出端到端脑-文本神经接口BIT，显著提升解码内心语音的准确率。	contrastive learning large language model
22	Cross-cultural value alignment frameworks for responsible AI governance: Evidence from China-West comparative analysis	提出多层审计平台，评估中西方LLM的跨文化价值对齐问题	reinforcement learning large language model
23	Improving Latent Reasoning in LLMs via Soft Concept Mixing	提出软概念混合（SCM）训练方案，提升LLM在抽象推理任务中的潜在推理能力	reinforcement learning large language model

🔬 支柱六：视频提取与匹配 (Video Extraction) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
24	HUMORCHAIN: Theory-Guided Multi-Stage Reasoning for Interpretable Multimodal Humor Generation	提出HUMORCHAIN，通过理论引导的多阶段推理生成可解释的多模态幽默内容	HuMoR large language model multimodal

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
25	Deep Improvement Supervision	提出深度改进监督方法，提升小型循环模型在复杂推理任务中的效率。	classifier-free guidance large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页