cs.CL（2026-04-10）

📊 共 19 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) (13) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱六：视频提取与匹配 (Video Extraction) (1 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (13 篇)

#	题目	一句话要点	标签	🔗	⭐
1	GRASP: Grounded CoT Reasoning with Dual-Stage Optimization for Multimodal Sarcasm Target Identification	提出GRASP框架，通过双阶段优化和Grounded CoT推理解决多模态讽刺目标识别问题。	multimodal visual grounding chain-of-thought
2	Hierarchical Alignment: Enforcing Hierarchical Instruction-Following in LLMs through Logical Consistency	提出神经符号分层对齐（NSHA），通过逻辑一致性增强LLM的分层指令遵循能力	large language model instruction following
3	Large Language Models Generate Harmful Content Using a Distinct, Unified Mechanism	通过权重剪枝揭示大语言模型生成有害内容的统一机制	large language model
4	CONDESION-BENCH: Conditional Decision-Making of Large Language Models in Compositional Action Space	提出CONDESION-BENCH，评估大语言模型在组合动作空间中的条件决策能力	large language model
5	Breaking Block Boundaries: Anchor-based History-stable Decoding for Diffusion Large Language Models	提出基于锚点的历史稳定解码AHD，解决扩散大语言模型中半自回归解码的块约束问题。	large language model
6	Many-Tier Instruction Hierarchy in LLM Agents	提出ManyIH以解决LLM Agent中多层级指令冲突问题，并构建相应评测基准。	large language model instruction following
7	Automated Instruction Revision (AIR): A Structured Comparison of Task Adaptation Strategies for LLM	提出AIR：一种基于规则归纳的LLM自动化指令修订方法，用于下游任务自适应。	large language model
8	BERT-as-a-Judge: A Robust Alternative to Lexical Methods for Efficient Reference-Based LLM Evaluation	提出BERT-as-a-Judge，用于高效、鲁棒的基于参考答案的LLM评估。	large language model
9	Persona-E$^2$: A Human-Grounded Dataset for Personality-Shaped Emotional Responses to Textual Events	提出Persona-E$^2$数据集，用于研究人格对文本事件情感反应的影响	large language model
10	MuTSE: A Human-in-the-Loop Multi-use Text Simplification Evaluator	MuTSE：人机协同的多用途文本简化评估器，解决LLM文本简化评估难题。	large language model
11	Across the Levels of Analysis: Explaining Predictive Processing in Humans Requires More Than Machine-Estimated Probabilities	批判性分析语言模型在人类语言处理中的作用，强调多层次分析的必要性	large language model
12	Task-Aware LLM Routing with Multi-Level Task-Profile-Guided Data Synthesis for Cold-Start Scenarios	提出TRouter，通过多级任务画像引导的数据合成，解决冷启动场景下的LLM路由问题。	large language model
13	TaxPraBen: A Scalable Benchmark for Structured Evaluation of LLMs in Chinese Real-World Tax Practice	TaxPraBen：构建中文税务实践LLM结构化评估基准	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	🔗	⭐
14	From Reasoning to Agentic: Credit Assignment in Reinforcement Learning for Large Language Models	针对大语言模型在强化学习中信用分配问题，提出系统性综述、评估基准与方法论指导。	reinforcement learning large language model chain-of-thought
15	Think Less, Know More: State-Aware Reasoning Compression with Knowledge Guidance for Efficient Reasoning	提出STACK框架，通过知识引导和状态感知压缩提升大语言模型推理效率。	PPO DPO direct preference optimization
16	Cross-Lingual Attention Distillation with Personality-Informed Generative Augmentation for Multilingual Personality Recognition	提出ADAM框架，通过跨语言注意力蒸馏和个性化生成增强实现多语言人格识别。	distillation large language model
17	PerMix-RLVR: Preserving Persona Expressivity under Verifiable-Reward Alignment	PerMix-RLVR：在可验证奖励对齐下保持角色表达能力	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	🔗	⭐
18	SPASM: Stable Persona-driven Agent Simulation for Multi-turn Dialogue Generation	SPASM：稳定角色驱动的Agent模拟框架，用于多轮对话生成	egocentric large language model	✅

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
19	Towards Linguistically-informed Representations for English as a Second or Foreign Language: Review, Construction and Application	构建面向ESFL的语言学知识表示，用于提升二语习得研究	semantic mapping semantic map

⬅️ 返回 cs.CL 首页 · 🏠 返回主页