cs.AI（2026-04-24）

📊 共 12 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) (8) 支柱二：RL算法与架构 (RL & Architecture) (4 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (8 篇)

#题目一句话要点标签🔗⭐
1 Introducing Background Temperature to Characterise Hidden Randomness in Large Language Models 引入背景温度以表征大型语言模型中的隐性随机性 large language model
2 Evaluating LLM-Based Goal Extraction in Requirements Engineering: Prompting Strategies and Their Limitations 提出基于LLM链和生成-评价机制的Goal-Oriented需求工程自动化方法。 large language model chain-of-thought
3 ResRank: Unifying Retrieval and Listwise Reranking via End-to-End Joint Training with Residual Passage Compression ResRank：通过残差通道压缩和端到端联合训练统一检索和列表式重排序 large language model multimodal
4 Superminds Test: Actively Evaluating Collective Intelligence of Agent Society via Probing Agents 提出Superminds Test，评估大规模Agent社会中的集体智能涌现现象 large language model
5 From Natural Language to Verified Code: Toward AI Assisted Problem-to-Code Generation with Dafny-Based Formal Verification 提出NL2VC-60数据集，结合Dafny形式验证，提升LLM代码生成正确率。 large language model
6 Rethinking Math Reasoning Evaluation: A Robust LLM-as-a-Judge Framework Beyond Symbolic Rigidity 提出基于LLM的数学推理评估框架，提升评估鲁棒性，超越符号刚性 large language model
7 SSG: Logit-Balanced Vocabulary Partitioning for LLM Watermarking 提出SSG：一种logit平衡的词汇划分方法，提升LLM水印在低熵场景下的检测能力 large language model
8 BLAST: Benchmarking LLMs with ASP-based Structured Testing BLAST：提出基于ASP的结构化测试基准，评估LLM在ASP代码生成中的准确性 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🔗⭐
9 Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond 提出Agentic World Modeling框架，旨在构建具备预测、模拟和演化能力的智能体环境模型。 reinforcement learning world model world models
10 A Co-Evolutionary Theory of Human-AI Coexistence: Mutualism, Governance, and Dynamics in Complex Societies 提出人类-AI共生演化理论，构建多层动态系统模型以实现稳定共存。 world model world models embodied AI
11 UniSonate: A Unified Model for Speech, Music, and Sound Effect Generation with Text Instructions UniSonate：一个统一的文本指令语音、音乐和音效生成模型 flow matching curriculum learning multimodal ✅
12 Aligning Dense Retrievers with LLM Utility via DistillationAligning Dense Retrievers with LLM Utility via Distillation 提出Utility-Aligned Embeddings (UAE)，通过蒸馏LLM效用对齐稠密检索器，提升检索性能。 distillation

⬅️ 返回 cs.AI 首页 · 🏠 返回主页