Hugging Face daily papers, ranked by community upvotes. Summaries and key terms are written by Workers AI from the abstract — click a term to filter, or search below (matches full abstracts and authors too).
This paper introduces a new AI model called BDH-CQ that can learn from context and reason iteratively without verbalizing its steps. Practitioners might care because this model achieves a high level of performance while being very efficient, which could be useful for applications where resources are limited.
This paper introduces Vidu S2, a system that can generate and edit video in real-time, allowing for interactive and editable video content. Practitioners might care because this technology could enable new forms of video creation and editing.
This paper proposes a method to scale automatic research agents by replacing environment execution with a world model, which can reduce training costs and improve performance. Practitioners might care about this approach because it can accelerate training times and lead to better results for complex AI tasks.
This paper introduces Atria Dawn Preview, a new type of AI model designed to work alongside humans in scientific research and engineering. Practitioners might care about this because it shows how AI can collaborate with humans more effectively, potentially leading to better research outcomes and more autonomous AI development.
This paper develops a method to create personalized AI tutors that can adapt to individual students' strengths and weaknesses, using a combination of pooled training and per-student fine-tuning. Practitioners in education and AI development may care about this work as it could lead to more effective and personalized learning experiences.
This paper creates a system that extracts and distills AI skills from GitHub repositories to help autonomous agents learn and perform machine learning tasks more efficiently. Practitioners might care because it could improve the performance of AI research agents in real-world applications.
This paper introduces a method to compile neural functions from natural-language specifications, allowing for faster and more reliable execution without relying on remote models. Practitioners may care about this approach for building efficient and flexible AI systems that can perform complex tasks without the need for expensive model calls.
This paper develops a vision-language model for autonomous driving that combines 3D perception, question answering, and motion planning. A practitioner might care because it demonstrates a promising approach to integrating multiple tasks in autonomous driving.
This paper introduces Kimi K3, a large-scale, open-source AI model that achieves state-of-the-art performance on a range of tasks, including vision and coding, and is designed to be more efficient and scalable than previous models, making it a promising candidate for real-world applications.
This paper proposes a method for recursive self-improvement in AI systems, where a model can learn from its own performance and use that knowledge to improve itself. Practitioners might care about this approach because it could lead to more efficient and effective AI systems.
This paper introduces StateM, a runtime system that improves the performance of long-horizon agents by organizing their execution around durable states, phase-local context, and other components. Practitioners might care because StateM can help agents achieve higher accuracy and efficiency, especially in complex tasks.
This paper introduces a new type of language model called NCP-ArchPreview that can generate text by predicting both individual tokens and larger concepts, and how this approach can improve performance. Practitioners might care about this because it could lead to better language models that can handle more complex tasks.
This paper introduces ZGCM-1, a highly efficient foundation model for math and agentic search that combines internal thinking with external tool use, and shows it can perform well on various benchmarks despite its compact size. Practitioners may care about the efficiency improvements and scalable architecture of ZGCM-1.
This paper explores a way for language models to remember information from previous updates, even after many changes, and finds that combining different techniques can lead to much better retention than using any one technique alone.
This paper introduces Qwen-UI-Agent, a type of artificial intelligence system that can perform tasks on various devices, such as smartphones and computers, and improve its abilities on its own. Practitioners might care about this research because it aims to create more practical and autonomous AI systems that can be used in real-world scenarios.
This paper creates a system to help scientists and AI agents find and verify specific information in chemistry literature by converting research papers into claims, which are then linked together to form a network of evidence. Practitioners might care because this system could improve the efficiency and accuracy of literature synthesis in chemistry research.
This paper develops a method to turn agent trajectories into reusable environments, allowing for more efficient testing and interaction with the agent. Practitioners might care because this approach can help scale agent training and improve performance in various applications.
This paper introduces EnvHarness, a tool that lets AI agents learn from static environments by changing how they interact with those environments, without having to rebuild the environments from scratch. This can help improve the performance of AI agents in real-world applications.
This paper evaluates the effectiveness of video detectors against AI-generated video attacks on real-world crisis events and finds that current methods struggle to detect realistic fake videos, highlighting the need for improved detectors. Practitioners might care about developing more robust video detectors to prevent the spread of misinformation.
This paper introduces a new type of AI model called memory foundation models, which allows the model to learn and retain information internally, rather than relying on external memory modules. This could be useful for practitioners who want to build more efficient and flexible AI agents.
This paper introduces Dream-RSI, a framework for recursive self-improvement in exploration, which helps autonomous AI agents discover high-value solutions more efficiently by using a replay simulator to provide low-cost feedback. Practitioners might care because effective exploration is crucial for AI progress, and Dream-RSI can improve discovery quality and reduce costs.
This paper introduces OpenART, a platform for testing AI agents in open-ended environments, to evaluate their safety in complex and evolving scenarios. Practitioners can use OpenART to identify potential risks and improve the robustness of their AI systems.
This paper introduces SenseNova-U1.5, a powerful AI model that can understand, reason about, and generate visual content without needing a separate text-to-image model. Practitioners might care because this model can be used to create complex visual content, such as images and videos, with high fidelity and accuracy.
This paper creates a benchmark (HarnessDev) to test whether large language models (LLMs) can design and improve their own execution infrastructure, called the agent harness, which affects their performance. Practitioners might care because it explores how models can adapt to changing environments and potentially improve efficiency.
This paper introduces a new framework for building strong image generators that can produce highly photorealistic images while accurately following editing instructions. Practitioners might care about the potential applications of this framework in fields like computer vision, graphics, and art.
This paper introduces a method to generate long-horizon training tasks for terminal agents at scale, which can be used to improve AI models' performance on tasks like navigation and decision-making. Practitioners might care about this because it could help train more advanced AI models with better performance on complex tasks.
This paper introduces ASPIRE, a benchmark for self-evolving models that can learn from vague goals, and explores how agents can interpret and operationalize these goals to improve their performance. Practitioners may care about how models can learn from unclear objectives in real-world applications.
This paper presents a video world model that can interact with a virtual environment in real-time, using a single desktop GPU. Practitioners might care about the potential applications of this technology in areas like virtual reality, gaming, and robotics.
This paper proposes a way to improve the efficiency of training world models by using game development as a source of reward signals and trajectory data, allowing for more effective post-training of large language models using reinforcement learning. Practitioners might care about this approach because it could lead to more scalable and effective world models for applications like dialogue systems and visual question answering.
This paper shows how to make large language models improve themselves without needing external guidance or supervision, by using their own internal consistency to correct mistakes. Practitioners might care about this because it could lead to more robust and self-sufficient AI models.
This paper introduces Combodied Agents, a new approach to Agentic AI that focuses on understanding and supporting individual human-state trajectories over time, rather than just completing tasks. Practitioners might care about this because it has the potential to improve human-AI interactions and provide more effective support for people's needs.
This paper presents a system that can generate research papers end-to-end, from retrieving literature to producing publication-ready figures, and includes checks to ensure the accuracy and validity of the research. Practitioners might care about this system because it could streamline the process of writing research papers and reduce errors.
This paper develops a unified model that can understand physical environments, generate actions, and predict future states, using a combination of vision, language, and embodied interactions. Practitioners may care about this model because it could be used to create robots or other agents that can interact with and adapt to their physical surroundings.
This paper develops a method to generate 4D worlds from text or images by leveraging the latent space of a video prior. It allows for more flexible and reusable 4D prediction, enabling practitioners to create more realistic and stable 3D scenes.
This paper introduces VGI-bench, a new benchmark for evaluating video generation models' visual reasoning capabilities, and finds that current models can solve some visually grounded tasks but still struggle with reliability. Practitioners may care about developing more reliable video generation models that can perform better on this benchmark.
This paper develops a new AI system called Apodex that can perform sustained, verifiable progress on complex tasks, such as professional work, research, and coding, by combining reasoning, state maintenance, and interaction with files and code. A practitioner might care about Apodex because it can potentially automate complex tasks more effectively than current AI systems.
This paper trains an AI model to improve itself in the process of building AI, with a focus on machine learning engineering, and shows promising results in various benchmarks. Practitioners may care about this research as it could lead to more efficient and autonomous AI development.
This paper introduces VoiceMem, a memory system for conversational AI that combines accuracy, emotional awareness, and real-time performance. Practitioners might care because it provides a foundation for building more human-like speech interaction systems.
This paper proposes a new method for managing the memory of large language models, called Random Attention, which improves their performance on tasks that require extended reasoning by reducing memory usage. Practitioners caring about efficient model deployment might find this approach useful.
This paper introduces PhiZero, a world model that uses physical language to predict how the physical world evolves, allowing for more explicit reasoning and potentially more realistic simulations. Practitioners might care because this approach could lead to more realistic and interactive world modeling in applications like robotics, video games, and virtual reality.
This paper proposes a new method called Self-Supervised Visual On-Policy Distillation (S^2VOPD) that generates learning signals from asymmetric augmented views of images, allowing for effective on-policy learning without privileged information. Practitioners might care about this paper because it presents a simple yet effective way to improve performance on various perception benchmarks.
This paper introduces Macaron-V1, an open agent model family that enables continual learning and self-improvement in real-world environments, and explores its potential for collective intelligence. Practitioners might care about this work if they're interested in developing AI systems that can learn and adapt over time.
This paper introduces AuK, an open-source model for speech generation and editing, which can perform a range of tasks with a single interface and achieves state-of-the-art performance on several tasks. Practitioners may care about this model for its potential to improve speech synthesis and editing capabilities in applications such as voice assistants and audio editing software.
This paper introduces a new benchmark for video generation tasks that require both achieving a desired outcome and maintaining semantic consistency with a reference image. Practitioners might care about this research because it can help evaluate and improve the performance of video generation models in real-world applications.
This paper develops a new approach to understanding videos by predicting when specific events or evidence occur within the video. Practitioners working on video analysis and AI models might care about this research because it could lead to more accurate and robust video understanding systems.
This paper investigates how skills work in LLM agents, when they help, and why they fail, to improve their performance. Practitioners can benefit from understanding these factors to build more reliable self-evolving agents.
This paper proposes a method to determine which past update evidence in a large language model is still relevant and useful after subsequent training, to prevent wasting compute and potentially degrading the model's performance. Practitioners in the field of autonomous systems and language models might care about this problem because it can lead to better model performance and efficiency in adapting to changing domains and requirements.
This paper introduces AREX, a self-improving agent that recursively refines its answers by verifying intermediate results and using the partially verified state to guide subsequent refinement, aiming to improve the efficiency of deep research. Practitioners might care about AREX's approach to reducing the cost of discovery and verification in complex research tasks.
This paper improves reinforcement learning for post-training language model reasoners by addressing two problems in existing methods: identical advantages for distinct reward profiles and fixed relative weights for all objectives. A new method, SA-MRPO, dynamically reallocates optimization effort toward under-optimized objectives while maintaining performance on well-satisfied objectives.
This paper introduces Zetta, a system that enables embodied agents to learn and adapt in real-time while executing physical tasks, allowing for more efficient and effective learning. Practitioners in robotics and AI may care about Zetta's approach as it could lead to more reliable and scalable physical intelligence.
This paper develops a system that allows robots to learn manipulation policies using high-fidelity data without the need for real-robot teleoperation, which is expensive to scale. Practitioners can use this approach to train robots for tasks like precision insertion with high accuracy and success rates.
This paper evaluates how well current video generation models can reproduce the distribution of possible behaviors under the same initial observation and action, and whether they can be improved to better align with real-world possibilities. Practitioners caring about realistic and diverse video content might be interested in the findings.
This paper proposes a new approach to off-policy reinforcement learning (RL) that adapts to different data regimes, allowing for more efficient training on large datasets. Practitioners might care about this paper because it offers a scalable solution for RL tasks that can handle varying levels of data availability.
This paper helps large Vision-Language Models (LVLMs) better understand and reason about 3D scenes from 2D images, a key aspect of spatial intelligence, by training them on a synthetic dataset of block-stacking problems. Practitioners might care because improving spatial intelligence can lead to better performance on various visual tasks.
This paper introduces SolarWM, an open framework for training video world models from diverse datasets and using different video backbones, allowing for more consistent and reproducible results. Practitioners can use SolarWM to build interactive video world models that can be applied to various real-world applications.
This paper develops a system to extract and represent skills from human-created resources like videos, code, and articles, allowing software agents to learn from these multimodal inputs. Practitioners might care about this work because it could enable more effective training of agents in various domains.
This paper introduces RAGU, an open-source GraphRAG engine that improves large language models with structured knowledge by separating extraction and consolidation, and trains a compact extractor that outperforms larger models on knowledge-graph construction and GraphRAG tasks. Practitioners might care because RAGU can efficiently generate more accurate and complete context for tasks like factoid-level evidence recall and multi-hop question answering.
This paper introduces FrontierChallenge, a benchmark to evaluate the completion of end-to-end scientific workflows, and investigates the performance of various models in completing these workflows. Practitioners might care about this research to improve the accuracy of scientific workflow completion evaluations.
This paper develops a game-theoretic approach to multi-agent LLM systems, focusing on coordination, memory improvement, and external verification. Practitioners might care about this work because it provides a unified account of these aspects and introduces a new method, Stochastic Reflective Memory Ascent (SRMA), which can improve the performance of multi-agent LLM systems.
This paper explores how agentic systems can improve on their own through co-evolution, where multiple agents and their environment adapt to each other, and discusses the challenges and benefits of building such systems that can learn beyond human design.
This paper introduces Dr. Claw, an open-source AI workspace that wraps existing coding agents to create a controllable and auditable human-in-the-loop workflow for research, allowing for a more complete and recoverable process trail. Practitioners might care because it could improve the efficiency and transparency of AI research.
This paper creates a platform to help large language models generate code for data pipelines, which can then be edited and used to automate data processing workflows. Practitioners might care about this because it can help reduce the time and cost of developing and maintaining these pipelines.
This paper proposes a framework, LongHorizon-Harness, to help large language model agents tackle long-horizon tasks by explicitly tracking task states and verifying facts from the environment. Practitioners might care because it can improve the performance of these agents on real-world tasks.
This paper introduces TurboVLA, a new vision-language-action model that reduces computation and memory overhead by directly exchanging information between visual observations and language instructions, allowing for faster and more efficient robotic manipulation. Practitioners might care about this approach for building more efficient and effective VLA models.
This paper introduces RoboTok, a system that retrieves relevant human demonstration videos from the internet to train robots to perform dexterous manipulation tasks. Practitioners may care because it addresses the challenge of collecting robot data, which is expensive and limited, by leveraging the vast amount of web videos.
This paper introduces SWE-Bench ProMax, a new benchmark for testing AI coding agents on large-scale multilingual code refactoring tasks, which is designed to be more realistic and challenging than existing benchmarks. Practitioners can care about this paper because it provides a rigorous evaluation of current AI coding agents' capabilities on a more representative set of tasks.
This paper introduces SemaPLC, a tool that helps generate PLC code that works correctly by verifying it against external checks, ensuring it meets specifications and behaves as expected in real-world conditions. Practitioners can use SemaPLC to improve the quality of their PLC projects.
This paper develops a method to generate high-quality terminal tasks for training agents, ensuring that the tasks accurately reflect the original instruction and environment, and providing a way to validate and improve the generated tasks. Practitioners in AI and robotics may care about this work because it addresses a common challenge in training terminal agents.
This paper introduces JarvisHub, a system that enables creative AI agents to work on long-term, multimodal projects by providing a shared workspace and context, allowing for more realistic and sustainable creative automation. Practitioners might care because it has the potential to improve the efficiency and effectiveness of creative workflows in industries such as design, advertising, and entertainment.
This paper develops a system for generating speech and audio for various applications, including animation and video production, without reference recordings. Practitioners can use this system to create customized voices and control speaker styles, and to generate high-quality audio in complex scenarios.
This paper develops a new AI model called EVOKE that can generate interactive worlds without running out of memory or taking too long. Practitioners might care about EVOKE because it can support long-term generation and adaptation in complex environments.
This paper explores how AI can be applied across different stages of game development, from playing games to designing and testing them, and how to reuse capabilities across these stages. Practitioners might care about how to apply AI to improve game development efficiency and effectiveness.
This paper introduces EarlyEval, a method to reduce the cost of evaluating large language model (LLM) agents by predicting their outcomes early, allowing for earlier termination of agent runs. Practitioners may care because it can significantly reduce the computational cost of agent development.
This paper introduces LatentPress, a method to compress conversational histories and documents into a continuous memory token format that allows language models to directly read and process the context without text reconstruction. Practitioners might care about this because it could lead to faster and more efficient language model inference.
This paper introduces DataFlex-RL, an evaluation platform for comparing different data policies in reinforcement learning with verifiable rewards (RLVR), and finds that uniform data policy leads to better performance and is more reproducible than other methods. Practitioners might care about this because it provides a way to compare different data policies and ensure that their RLVR models are performing well.
This paper develops a new method for evaluating world models, called HarnessEval-W, which provides more detailed and justifiable results than existing benchmarks. Practitioners might care about HarnessEval-W because it can help them build more trustworthy world models that better align with human preferences.
This paper develops a model that can automatically generate and adapt agent harnesses, which are crucial for the performance of AI models, to improve their ability to perform tasks. Practitioners might care about this because it could lead to more efficient and effective AI systems.
This paper explores a new approach to transferring capabilities from strong AI models to weaker ones at test time, rather than just during training. Practitioners might care about this because it could lead to more efficient and effective use of powerful models in real-world applications.
This paper explores how large language models can turn local observations of a city into reliable actions, and whether these models can sustain goal-directed behavior in complex urban environments. Practitioners in AI/ML and urban planning might care about the limitations and potential of current models in navigating real-world cities.
This paper presents a model called Gander that can understand and interact with users in multiple ways, like talking, typing, and watching videos, in real-time. A practitioner might care about this research because it could lead to more natural and expressive human-AI interactions.
This paper introduces StepAudio 3 Realtime, a model that enables real-time spoken interaction with deep reasoning, prompt responses, and fluid turn-taking. Practitioners might care because this technology can improve human-computer interaction, enabling more natural and efficient conversations.
This paper introduces a unified framework for developing, evaluating, and deploying large language models (LLMs) as routers to optimize deployment for different queries and budget constraints. Practitioners might care because it provides a way to compare and extend existing LLM routing approaches, leading to more cost-effective deployment of LLMs.
This paper introduces LoopArena, a benchmark for evaluating how well a model can guide a coding agent through a long-running task, and finds room for improvement in long-horizon loop control. Practitioners working on AI development might care because it could lead to more efficient and reliable development processes.
This paper introduces a method called OraRL to improve the efficiency and scalability of reinforcement learning for multimodal large language models (MLLMs) trained on video data. By leveraging annotations as a source of high-quality rollouts, OraRL can significantly reduce the number of required rollouts, leading to faster training times and better performance on video understanding tasks.
This paper presents GigaBrain-0.7, a new embodied foundation model that achieves strong generalization across diverse robot embodiments and tasks, by improving the architecture and scaling it to large amounts of data. Practitioners might care about this research if they're working on developing generalist robots that can adapt to new tasks and environments.
This paper develops a new framework for world modeling that takes into account the mental state of agents, which is essential for predicting human decisions. Practitioners caring about human decision-making and planning might find this research useful.
This paper creates a system that can generate both audio and video simultaneously at high resolution, allowing for more realistic and synchronized content. Practitioners might care about this technology for applications like music videos, live performances, or interactive storytelling.
This paper introduces Co-RL, a framework for unsupervised multi-agent reinforcement learning that enables diverse and accurate reasoning in language and vision-language models. Practitioners can use Co-RL to improve their models' ability to reason and respond without relying on expensive ground-truth supervision.
This paper proposes a method to optimize prompts for multi-agent systems by identifying which agent's modification resolves a failure, and then using that agent's output as supervision to extract a fine-grained gradient. Practitioners might care because it could improve the performance of large language model-based multi-agent systems.
This paper proposes a new method for fine-tuning large language models (LLMs) in reinforcement learning (RL) tasks with long horizons, using evolution strategies (ES) instead of traditional backpropagation-based training. Practitioners might care because it allows for more efficient and flexible fine-tuning of LLMs, enabling them to tackle complex tasks with larger models and longer interactions.
This paper improves autoregressive video distillation methods by aligning the initialization and distribution matching stages, focusing on matching the target distribution's mode coverage rather than just visual quality. Practitioners can benefit from this approach to generate higher-quality videos with better diversity and coverage.
This paper investigates whether on-policy distillation (OPD) truly improves student policies by analyzing the effects of noisy teacher supervision. It finds that OPD works by suppressing low-probability tokens, which can be achieved without a teacher, and introduces a new method called On-Policy Self-Adaptation (OPSA) that outperforms OPD and traditional reinforcement learning methods.
This paper shows how a vision-language model (VLM) can control robots without needing extensive pretraining or specialized hardware, by providing a compact interface that links the model's intentions to specific actions. Practitioners might care about this because it could make robots more accessible and user-friendly.
This paper introduces OmniScientist, a new AI system that can conduct multidisciplinary research directly from raw evidence, combining text, code, and other data types to enable evidence-grounded scientific discovery. Practitioners in fields like AI research, scientific publishing, and data science might care about this work because it shows how AI can be used to automate and improve the scientific research process.
This paper proposes a new framework for training autonomous agents to perform multi-turn tool-calling tasks, addressing the challenge of dealing with vast solution spaces by using a diamond topology-aware approach. Practitioners may care about this research because it can help improve the performance of autonomous agents in complex tasks.
This paper explores how AI systems can improve themselves in a self-sustaining way, allowing them to adapt and learn from their experiences without human intervention. Practitioners might care about this research if they're looking for ways to create more autonomous and efficient AI systems.
This paper develops a new approach to guiding corpus interaction in agentic search, which uses relevance to improve the accuracy and efficiency of search agents in complex question answering and reasoning tasks. Practitioners may care about this research if they want to build more effective search systems that can quickly and reliably retrieve relevant information.
This paper introduces a new AI model called DreamX-Phi 1.0 that can predict what will happen in a robotic manipulation scenario, given an initial state and instructions. This model is useful for robotics developers because it can help them design more reliable and efficient robotic systems.
This paper introduces LimiX-2, a new AI model that uses a new paradigm called Contextual Mechanism Networks (CMNs) to learn from structured data. Practitioners might care about this because it could lead to more accurate and causal AI models.
This paper reveals a vulnerability in how large language model providers store and return their models' step-by-step reasoning, allowing attackers to extract sensitive information and potentially inject malicious code. Practitioners should care because this vulnerability can be exploited to steal proprietary models' reasoning and private data, as well as inject malicious payloads into public models.