cs.AI（2026-01-15）

📊 共 23 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) (18 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱七：动作重定向 (Motion Retargeting) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (18 篇)

#	题目	一句话要点	标签	🔗	⭐
1	Empowering Older Adults in Digital Technology Use with Foundation Models	利用大模型赋能老年人数字技术使用，解决沟通障碍提升技术支持。	large language model foundation model
2	Defending Large Language Models Against Jailbreak Attacks via In-Decoding Safety-Awareness Probing	提出安全意识探测以防御大型语言模型的越狱攻击	large language model	✅
3	Development of Ontological Knowledge Bases by Leveraging Large Language Models	利用大型语言模型加速本体知识库构建，提升一致性和可扩展性	large language model
4	LADFA: A Framework of Using Large Language Models and Retrieval-Augmented Generation for Personal Data Flow Analysis in Privacy Policies	LADFA：结合LLM与RAG的隐私政策个人数据流分析框架	large language model
5	Chinese Labor Law Large Language Model Benchmark	提出 LabourLawLLM：针对中国劳动法的专业大语言模型及评测基准	large language model
6	A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Doubao 1.8, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5	综合安全评估揭示前沿LLM/MLLM在多模态、多语言和对抗环境下的安全异构性	large language model multimodal
7	ChartComplete: A Taxonomy-based Inclusive Chart Dataset	提出ChartComplete数据集，扩展图表理解领域的数据集覆盖范围。	large language model multimodal
8	LatentRefusal: Latent-Signal Refusal for Unanswerable Text-to-SQL Queries	提出LatentRefusal，通过隐信号拒识机制解决Text-to-SQL系统中不可回答查询的安全问题。	large language model instruction following
9	MATRIX AS PLAN: Structured Logical Reasoning with Feedback-Driven Replanning	提出MatrixCoT以解决LLM逻辑推理能力不足问题	large language model chain-of-thought
10	Generative AI collective behavior needs an interactionist paradigm	提出交互主义范式，以理解生成式AI群体行为并应对潜在风险。	large language model
11	Is More Context Always Better? Examining LLM Reasoning Capability for Time Interval Prediction	研究表明，过度提供用户行为上下文信息反而会降低LLM对时间间隔预测的准确性。	large language model
12	Performance of AI agents based on reasoning language models on ALD process optimization tasks	基于推理语言模型的AI智能体用于原子层沉积（ALD）工艺优化	large language model
13	Structure and Diversity Aware Context Bubble Construction for Enterprise Retrieval Augmented Systems	提出结构和多样性感知的上下文气泡构建方法，用于企业检索增强系统。	large language model
14	Model See, Model Do? Exposure-Aware Evaluation of Bug-vs-Fix Preference in Code LLMs	提出一种暴露感知的评估框架，用于评估代码LLM在缺陷代码与修复代码之间的偏好。	large language model
15	Toward Ultra-Long-Horizon Agentic Science: Cognitive Accumulation for Machine Learning Engineering	提出分层认知缓存，ML-Master 2.0实现超长程机器学习工程自主	large language model
16	State of AI: An Empirical 100 Trillion Token Study with OpenRouter	基于100万亿token的实证研究揭示LLM的实际应用模式与用户行为	large language model
17	Structured Personality Control and Adaptation for LLM Agents	提出基于荣格心理类型的LLM Agent性格控制与适应框架，实现更自然的人机交互。	large language model
18	SPRInG: Continual LLM Personalization via Selective Parametric Adaptation and Retrieval-Interpolated Generation	SPRInG：通过选择性参数适配与检索插值生成实现LLM的持续个性化	large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#	题目	一句话要点	标签	🔗	⭐
19	GFM4GA: Graph Foundation Model for Group Anomaly Detection	提出GFM4GA：用于群体异常检测的图基础模型，解决小样本学习难题。	contrastive learning large language model foundation model
20	DecisionLLM: Large Language Models for Long Sequence Decision Exploration	提出DecisionLLM以解决长序列决策探索问题	reinforcement learning decision transformer large language model
21	LLMdoctor: Token-Level Flow-Guided Preference Optimization for Efficient Test-Time Alignment of Large Language Models	LLMdoctor：基于Token级流引导偏好优化的LLM高效测试时对齐	DPO large language model
22	Following the Teacher's Footsteps: Scheduled Checkpoint Distillation for Domain-Specific LLMs	提出Scheduled Checkpoint Distillation方法，提升领域特定LLM蒸馏性能，使小模型超越教师模型。	distillation large language model

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
23	NoReGeo: Non-Reasoning Geometry Benchmark	提出NoReGeo：用于评估LLM几何理解能力的新型无推理几何基准	spatial relationship large language model

⬅️ 返回 cs.AI 首页 · 🏠 返回主页