H2025.02.07 | 特征流提升模型可解释性,超IF增强指令跟随能力。

2025.02.07 | 特征流提升模型可解释性,超IF增强指令跟随能力。

HuggingFace 每日AI论文速递·2025 年 2 月 8 日·15 分钟
在小宇宙播放
创作者写的节目说明

本期的 21 篇论文如下:

[] 🔄 Analyze Feature Flow to Enhance Interpretation and Steering in Language Models(分析特征流以增强语言模型的解释与控制)

[] 🤖 UltraIF: Advancing Instruction Following from the Wild(超IF:从野外推进指令跟随)

[] 🎥 DynVFX: Augmenting Real Videos with Dynamic Content(DynVFX:用动态内容增强真实视频)

[] 🌐 Ola: Pushing the Frontiers of Omni-Modal Language Model with Progressive Modality Alignment(Ola:通过渐进式模态对齐推动全模态语言模型的前沿)

[] 🏃 MotionLab: Unified Human Motion Generation and Editing via the Motion-Condition-Motion Paradigm(MotionLab:基于运动-条件-运动范式的统一人体运动生成与编辑)

[] 🤖 Great Models Think Alike and this Undermines AI Oversight(伟大的模型思维相似,这削弱了AI监督)

[] 📚 MAGA: MAssive Genre-Audience Reformulation to Pretraining Corpus Expansion(MAGA:大规模体裁-受众重构以扩展预训练语料库)

[] 🏆 Gold-medalist Performance in Solving Olympiad Geometry with AlphaGeometry2(在解决奥林匹克几何问题中实现金牌选手水平的AlphaGeometry2)

[] 🤖 ScoreFlow: Mastering LLM Agent Workflows via Score-based Preference Optimization(ScoreFlow:基于评分偏好优化的LLM代理工作流掌握)

[] 🎙 Llasa: Scaling Train-Time and Inference-Time Compute for Llama-based Speech Synthesis(Llasa:扩展基于Llama的语音合成中的训练和推理计算)

[] 🎥 MotionCanvas: Cinematic Shot Design with Controllable Image-to-Video Generation(MotionCanvas:基于可控图像到视频生成的电影镜头设计)

[] 📊 ChartCitor: Multi-Agent Framework for Fine-Grained Chart Visual Attribution(ChartCitor:细粒度图表视觉归属的多代理框架)

[] 🧠 BOLT: Bootstrap Long Chain-of-Thought in Language Models without Distillation(BOLT:无需蒸馏的大语言模型长链思维自举)

[] 🔄 Beyond Prompt Content: Enhancing LLM Performance via Content-Format Integrated Prompt Optimization(超越提示内容:通过内容-格式集成提示优化提升大语言模型性能)

[] 🌀 Weak-to-Strong Diffusion with Reflection(从弱到强扩散与反射)

[] 🤖 PlotGen: Multi-Agent LLM-based Scientific Data Visualization via Multimodal Feedback(PlotGen:基于多智能体LLM的科学数据可视化通过多模态反馈)

[] 🔧 Enhancing Code Generation for Low-Resource Languages: No Silver Bullet(提升低资源语言的代码生成:没有银弹)

[] 🔓 Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions(轻松对话:通过简单互动从LLM中引出有害越狱行为)

[] 🤖 PILAF: Optimal Human Preference Sampling for Reward Modeling(PILAF:最优人类偏好采样用于奖励建模)

[] 🎥 Towards Physical Understanding in Video Generation: A 3D Point Regularization Approach(面向视频生成的物理理解:一种3D点正则化方法)

[] 🤖 Learning Real-World Action-Video Dynamics with Heterogeneous Masked Autoregression(基于异质掩码自回归的现实世界动作视频动态学习)

【关注我们】

您还可以在以下平台找到我们,获得播客内容以外更多信息

小红书: AI速递


在小宇宙查看该单集文稿

说明里标出的时间点,点一下跳过去

HuggingFace 每日AI论文速递 · 更多单集

看全部 739 期 →

HuggingFace 每日AI论文速递

H2025.02.07 | 特征流提升模型可解释性,超IF增强指令跟随能力。

HuggingFace 每日AI论文速递 · 2025 年 2 月 8 日 · 15 分钟

2025.02.07 | 特征流提升模型可解释性,超IF增强指令跟随能力。

在小宇宙播放
创作者写的节目说明

本期的 21 篇论文如下:

[] 🔄 Analyze Feature Flow to Enhance Interpretation and Steering in Language Models(分析特征流以增强语言模型的解释与控制)

[] 🤖 UltraIF: Advancing Instruction Following from the Wild(超IF:从野外推进指令跟随)

[] 🎥 DynVFX: Augmenting Real Videos with Dynamic Content(DynVFX:用动态内容增强真实视频)

[] 🌐 Ola: Pushing the Frontiers of Omni-Modal Language Model with Progressive Modality Alignment(Ola:通过渐进式模态对齐推动全模态语言模型的前沿)

[] 🏃 MotionLab: Unified Human Motion Generation and Editing via the Motion-Condition-Motion Paradigm(MotionLab:基于运动-条件-运动范式的统一人体运动生成与编辑)

[] 🤖 Great Models Think Alike and this Undermines AI Oversight(伟大的模型思维相似,这削弱了AI监督)

[] 📚 MAGA: MAssive Genre-Audience Reformulation to Pretraining Corpus Expansion(MAGA:大规模体裁-受众重构以扩展预训练语料库)

[] 🏆 Gold-medalist Performance in Solving Olympiad Geometry with AlphaGeometry2(在解决奥林匹克几何问题中实现金牌选手水平的AlphaGeometry2)

[] 🤖 ScoreFlow: Mastering LLM Agent Workflows via Score-based Preference Optimization(ScoreFlow:基于评分偏好优化的LLM代理工作流掌握)

[] 🎙 Llasa: Scaling Train-Time and Inference-Time Compute for Llama-based Speech Synthesis(Llasa:扩展基于Llama的语音合成中的训练和推理计算)

[] 🎥 MotionCanvas: Cinematic Shot Design with Controllable Image-to-Video Generation(MotionCanvas:基于可控图像到视频生成的电影镜头设计)

[] 📊 ChartCitor: Multi-Agent Framework for Fine-Grained Chart Visual Attribution(ChartCitor:细粒度图表视觉归属的多代理框架)

[] 🧠 BOLT: Bootstrap Long Chain-of-Thought in Language Models without Distillation(BOLT:无需蒸馏的大语言模型长链思维自举)

[] 🔄 Beyond Prompt Content: Enhancing LLM Performance via Content-Format Integrated Prompt Optimization(超越提示内容:通过内容-格式集成提示优化提升大语言模型性能)

[] 🌀 Weak-to-Strong Diffusion with Reflection(从弱到强扩散与反射)

[] 🤖 PlotGen: Multi-Agent LLM-based Scientific Data Visualization via Multimodal Feedback(PlotGen:基于多智能体LLM的科学数据可视化通过多模态反馈)

[] 🔧 Enhancing Code Generation for Low-Resource Languages: No Silver Bullet(提升低资源语言的代码生成:没有银弹)

[] 🔓 Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions(轻松对话:通过简单互动从LLM中引出有害越狱行为)

[] 🤖 PILAF: Optimal Human Preference Sampling for Reward Modeling(PILAF:最优人类偏好采样用于奖励建模)

[] 🎥 Towards Physical Understanding in Video Generation: A 3D Point Regularization Approach(面向视频生成的物理理解:一种3D点正则化方法)

[] 🤖 Learning Real-World Action-Video Dynamics with Heterogeneous Masked Autoregression(基于异质掩码自回归的现实世界动作视频动态学习)

【关注我们】

您还可以在以下平台找到我们,获得播客内容以外更多信息

小红书: AI速递


在小宇宙查看该单集文稿

说明里标出的时间点,点一下跳过去

HHuggingFace 每日AI论文速递

HuggingFace 每日AI论文速递

面向中文AI从业者与爱好者的每日AI论文速递,工作日高密度速览前沿论文,月末以月度热门论文特辑补充,形成每日速递加月末特…

已收录 739 期 · 最近更新 11 月 10 日

在哪里听

https://feed.xyzfm.space/7jtt3j3lbbra复制

内容版权归创作者,声湃按创作者开放的 RSS 收录。了解收录与 RSS →

更多单集 全部 739 期 →