cs.CL（2026-04-16）

📊 共 18 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) (16 🔗3) 支柱一：机器人控制 (Robot Control) (1 🔗1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (16 篇)

#	题目	一句话要点	标签	🔗	⭐
1	Knowing When Not to Answer: Evaluating Abstention in Multimodal Reasoning Systems	提出MM-AQA基准，评估多模态推理系统中有效拒绝回答的能力	multimodal
2	IUQ: Interrogative Uncertainty Quantification for Long-Form Large Language Model Generation	提出IUQ框架，通过提问式不确定性量化提升长文本大语言模型生成结果的可信度。	large language model	✅
3	Compressing Sequences in the Latent Embedding Space: $K$-Token Merging for Large Language Models	提出K-Token Merging，通过潜在空间压缩减少LLM长文本处理的计算成本。	large language model
4	QuantCode-Bench: A Benchmark for Evaluating the Ability of Large Language Models to Generate Executable Algorithmic Trading Strategies	提出QuantCode-Bench基准，评估大语言模型生成可执行量化交易策略的能力	large language model
5	Blinded Multi-Rater Comparative Evaluation of a Large Language Model and Clinician-Authored Responses in CGM-Informed Diabetes Counseling	检索增强LLM在CGM指导的糖尿病咨询中表现优于临床医生	large language model
6	SPAGBias: Uncovering and Tracing Structured Spatial Gender Bias in Large Language Models	提出SPAGBias框架，揭示并追踪大语言模型中结构化的空间性别偏见	large language model
7	Reasoning Dynamics and the Limits of Monitoring Modality Reliance in Vision-Language Models	分析视觉语言模型推理动态，揭示其对不同模态依赖的局限性	multimodal chain-of-thought
8	CURaTE: Continual Unlearning in Real Time with Ensured Preservation of LLM Knowledge	CURaTE：提出一种支持实时持续卸载并保证LLM知识保留的框架	large language model
9	RACER: Retrieval-Augmented Contextual Rapid Speculative Decoding	提出RACER，融合检索与logits信息加速LLM推断，无需训练。	large language model	✅
10	Schema Key Wording as an Instruction Channel in Structured Generation under Constrained Decoding	提出将Schema关键词作为指令通道，提升约束解码下结构化生成性能。	large language model
11	CausalDetox: Causal Head Selection and Intervention for Language Model Detoxification	CausalDetox：通过因果头选择与干预实现语言模型解毒	large language model
12	From Tokens to Steps: Verification-Aware Speculative Decoding for Efficient Multi-Step Reasoning	提出SpecGuard，通过内部信号进行验证感知推测解码，提升多步推理效率。	large language model
13	Segment-Level Coherence for Robust Harmful Intent Probing in LLMs	提出基于片段一致性的流式探针，提升LLM在CBRN领域恶意意图检测的鲁棒性。	large language model
14	Modeling LLM Unlearning as an Asymmetric Two-Task Learning Problem	将LLM非学习建模为非对称双任务学习问题，提升知识遗忘与能力保持的平衡	large language model
15	Pushing the Boundaries of Multiple Choice Evaluation to One Hundred Options	提出大规模选项评估方法，用于更可靠地评估语言模型在密集干扰下的性能。	large language model
16	PeerPrism: Peer Evaluation Expertise vs Review-writing AI	提出PeerPrism基准，用于评估同行评审中人类专家与AI写作的贡献区分。	large language model	✅

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
17	Fact4ac at the Financial Misinformation Detection Challenge Task: Reference-Free Financial Misinformation Detection via Fine-Tuning and Few-Shot Prompting of Large Language Models	提出基于微调和少样本提示的大语言模型金融虚假信息检测方法，无需外部参考。	manipulation large language model	✅

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
18	ClimateCause: Complex and Implicit Causal Structures in Climate Reports	ClimateCause：构建气候报告因果关系数据集，用于提升模型气候变化理解能力	spatiotemporal large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页