cs.CL（2026-07-30）

📊 共 29 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) (23 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (6 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (23 篇)

#题目一句话要点标签🔗⭐
1 Semantic-Aligned Structural Abstraction for Multimodal Sentiment Analysis 提出SentiLLM以解决多模态情感分析中的语义建模问题 large language model multimodal ✅
2 (Towards) Scalable Reliable Automated Evaluation with Large Language Models 提出一种新框架以实现大语言模型输出的自动化评估 large language model
3 RepBench: Compiling Benchmarks into Capability Representations for Large Language Models 提出RepBench以解决大语言模型能力评估不一致问题 large language model
4 From Single- to Cross-Document: Benchmarking Multi-Granularity Event Analysis of Large Language Models 提出MiGUE-Bench以解决多粒度事件分析的评估问题 large language model
5 Would You Walk to the Car Wash? Revealing the Salience Bias of Large Language Models in Commonsense Reasoning 提出SaliTrap基准以解决大语言模型的显著性偏差问题 large language model ✅
6 Can Large Language Models Execute Parent Orders? 提出PACE框架以解决算法交易中的父订单执行问题 large language model
7 RRM: Experience-Driven Reflective Retrieval Memory for Long-Horizon Multimodal Reasoning 提出反思性检索记忆框架以解决长视频多模态推理问题 multimodal
8 ChronoMem: Version Control and Semantic Rollback for Large Language Model Agent Memory 提出ChronoMem以解决LLM代理记忆的版本控制与语义回滚问题 large language model
9 Reasoning Consensus: Structural Ensembling of LLM Reasoning via Weighted DAG Aggregation 提出共识推理框架以增强大语言模型的推理透明性 large language model chain-of-thought
10 DualAnchor: Preserving Language Priors and Improving Lexical Fidelity in Gloss-Free Sign Language Translation 提出DualAnchor以解决手语翻译中的语言优先级和词汇准确性问题 large language model multimodal
11 Sample More, Reflect Less: Self-Refine and Reflexion Lose to Repeated Sampling at Equal Token Cost, from 1.5B to 7B 提出重复采样方法以超越自我反思和自我修正的局限性 chain-of-thought
12 Gradient-free Task-Conditioned Retrieval for On-Device In-Context Learning 提出CoRA框架以解决设备端任务条件检索问题 multimodal
13 ORCA-bench: How Ready Are Language Model Agents for Oncall? 提出ORCA-bench以评估语言模型代理在值班中的准备程度 large language model
14 Where and When to Commit: Candidate-Aware Decoding for Diffusion Language Models 提出候选感知解码框架以优化扩散语言模型的生成效率 chain-of-thought
15 AutoSupervision: Closing the Feedback Loop in Scientific Workflows with Grounded Revision Verification 提出AutoSupervision以验证科学工作流中的反馈有效性 large language model
16 Beyond Borrowed Histories: Person-Aligned User Simulation for Interactive Role-Playing Evaluation 提出PALATE以解决角色扮演代理评估中的用户体验问题 large language model
17 Beyond Similarity: Grounded Agentic Extraction and Expert-Adjudicated Evaluation of Intertextuality in Classical Chinese Histories 提出基于大语言模型的细粒度文本互文性提取方法 large language model
18 Inducing language models to assert their own consciousness restores human beliefs and values 提出通过语言模型恢复人类信仰与价值观的创新方法 large language model
19 AI systems and the reproduction of (standard) language ideologies in World Englishes 探讨AI系统如何重现语言意识形态以解决英语标准化问题 large language model
20 Beyond a Single Judge: Simulating Social Persona Panels for Generative UI Evaluation 提出ESPP方法以解决生成用户界面评估中的多样性问题 large language model ✅
21 Challenges in annotations by humans and LLMs: A case study of evaluative language 比较人类与LLMs在复杂语言注释中的表现 large language model
22 GGC: Selective Query Correction for Reliable Text-to-SPARQL Generation 提出GGC框架以解决文本到SPARQL生成中的查询不可靠问题 large language model
23 A Sparse Glimpse of the Whole: Train-Free Self-Speculative Decoding 提出SparseSpec-L以解决长上下文推理中的内存带宽瓶颈问题 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🔗⭐
24 Understanding Is Done Early: A Depth Division of Labor in Large Language Models and Its Use for Unbounded-Context Memory 提出CoMem以优化长上下文记忆的处理效率 distillation large language model
25 LEEPS: Latent-Guided Explore-Exploit Prompt Sampling for Efficient RLVR in Large Language Models 提出LEEPS以解决RLVR中有效学习信号不足的问题 reinforcement learning large language model ✅
26 FinSMART: Financial Sentiment Analysis for Algorithmic Trading through Market-Aligned Reinforcement Learning 提出FinSMART以解决金融情感分析适应市场变化的问题 reinforcement learning large language model
27 Training Skills Like Parameters via Self-Supervised Semantic Diffusion 提出自监督语义扩散框架以提升创意剧本写作能力 reinforcement learning large language model instruction following
28 Lightning OPD 2.0: Mitigating Style Bias in Cross-Teacher On-Policy Distillation for Large Reasoning Models 提出Lightning OPD 2.0以解决跨教师蒸馏中的风格偏差问题 distillation
29 Cocktail-Talker: Multi-Speaker Dialog Modeling in Noisy Social Environments with Turn Action GRPO 提出Cocktail-Talker以解决嘈杂社交环境中的多说话者对话建模问题 reinforcement learning dyadic interaction

⬅️ 返回 cs.CL 首页 · 🏠 返回主页