Open library
The essential reading of generative AI and agents, open in a clean reading view, with Meow and Standard summaries when available. No account needed to read.
560 papers · 558 with a Meow summary · 333 posts
Every paper prepared on Reading + AI, newest first.
NVIDIA
**Abstract.** We introduce Nemotron 3 Ultra, a 550 billion total and 55 billion active parameter Mixture-of-Experts Hybrid Mamba-Attention language model. We pre-trained Nemotron 3 Ultra on 20 trillion text tokens, then extended the context length to 1M tokens, and post-trained using Supervised Fine Tuning (SFT), Rein…
NVIDIA
**Abstract. We introduce Nemotron 3 Nano Omni, the latest model in the Nemotron multimodal series and the first to natively support audio inputs alongside text, images, and video. Nemotron 3 Nano Omni delivers consistent accuracy improvements over its predecessor, Nemotron Nano V2 VL, across all modalities, enabled by…
Yonggan Fu, Lexington Whalen, Abhinav Garg et al.
We introduce Nemotron-Labs-Diffusion, a tri-mode language model (LM) that unifies AR, diffusion, and self-speculation decoding within a single architecture. Trained with a joint AR-diffusion objective, Nemotron-Labs-Diffusion can switch modes to sustain high throughput across deployment settings and concurrency levels…
Ivan Moshkov, Stephen Ge, George Armstrong et al.
**Abstract. We study how model post-training and test-time inference design affect natural-language proof generation for hard olympiad mathematics. Starting from Nemotron 3 Ultra, we train two specialist checkpoints using supervised fine-tuning and reinforcement learning, and evaluate checkpoint choice, verification,…
We present a full-pipeline inference optimization for the MiMo-V2.5 model family, which combines Hybrid Sliding Window Attention (Hybrid SWA), sparse Mixture-of-Experts (MoE), and multimodal encoders. While Hybrid SWA can ideally reduce both attention compute and KVCache storage significantly compared to Full Attentio…
LLM-Core Xiaomi
Reinforcement learning (RL) is the central training paradigm for advancing large foundation models towards self-improvement. This report introduces the MiMo-V2.6 series, an omni-modal family that pushes the frontier of model intelligence by scaling RL compute. Prior to RL, we conduct mid-training on a broad multimodal…
Xunhao Lai, Weiqi Xu, Yufeng Yang et al.
Ultra-long-context capability is becoming indispensable for frontier LLMs: agentic workflows, repository-scale code reasoning, and persistent memory all require the model to jointly attend over hundreds of thousands to millions of tokens—yet the quadratic cost of softmax attention makes this untenable at deployment sc…
Kimi Team
We introduce Kimi K3, a 2.8T parameter Mixture-of-Experts model with 104 billion activated parameters, native vision capabilities, and a 1-million-token context window. Kimi K3 is built on Kimi Delta Attention [[57](#bib.bib115)] and Attention Residuals [[60](#bib.bib56)], which improve information flow across sequenc…
We introduce the Ministral 3 series, a family of parameter-efficient dense language models designed for compute and memory constrained applications, available in three model sizes: 3B, 8B, and 14B parameters. For each model size, we release three variants: a pretrained base model for general-purpose use, an instructio…
Mingxin Li, Yanzhao Zhang, Dingkun Long et al.
In this report, we introduce the Qwen3-VL-Embedding and Qwen3-VL-Reranker model series, the latest extensions of the Qwen family built on the Qwen3-VL foundation model. Together, they provide an end-to-end pipeline for high-precision multimodal search by mapping diverse modalities, including text, images, document ima…
Qwen Team
In this report, we introduce Qwen3-ASR family, which includes two powerful all-in-one speech recognition models and a novel non-autoregressive speech forced alignment model. Qwen3-ASR-1.7B and Qwen3-ASR-0.6B are ASR models that support language identification and ASR for 52 languages and dialects. Both of them leverag…
The paper introduces GUI-Owl-1.5, the latest native GUI agent model that features instruct/thinking variants in multiple sizes (2B/4B/8B/32B/235B) and supports a range of platforms (desktop, mobile, browser, and more) to enable cloud-edge collaboration and real-time interaction. GUI-Owl-1.5 achieves state-of-the-art r…
MAI-UI Team
GUI agents have the potential to become a general purpose executor over existing digital devices. To advance them toward real-world use, we envision agents that operate reliably on real devices, execute workflows across platforms, combine GUI interaction with CLI execution, complete long-horizon tasks, proactively ini…
Qwen Team
We present **Qwen-Image-2.0**, an omni-capable image generation foundation model that unifies high-fidelity image generation and precise image editing within a single integrated framework. While current image generation foundation models excel at high-quality aesthetic generation and text rendering, they still face si…
Qwen Team
We present Qwen3-Coder-Next, an open-weight language model specialized for coding agents. Qwen3-Coder-Next is an 80-billion-parameter model that activates only 3 billion parameters during inference, enabling strong coding capability with efficient inference. In this work, we explore how far strong training recipes can…
Qwen Team
A world model predicts environment dynamics based on current observations and actions, serving as a core cognitive mechanism for reasoning and planning. In this work, we investigate how world modeling based on language models can further push the boundaries of general agents. (i) We first focus on building foundation…
Qwen Team
We describe the architecture and ablations of **Qwen3.8-Flash-Next**, a sparse mixture-of-experts model with 125B parameters, 6B activated per token, and additional 51B parameters of n-gram embedding tables held off the accelerator. On fourteen pre-training benchmarks the model leads the 397B-A17B predecessor on eight…
Shengbang Tong, David Fan, John Nguyen et al.
The visual world offers a critical axis for advancing foundation models beyond language. Despite growing interest in this direction, the design space for native multimodal models remains opaque. We provide empirical clarity through controlled, from-scratch pretraining experiments, isolating the factors that govern mul…
Qwen Team
We introduce Qwen3.8-Omni-Flash, a natively multimodal agentic model designed for real-world multimodal productivity. Compared with previous omni models, which primarily emphasized perception and interaction, Qwen3.8-Omni-Flash substantially improves multimodal understanding and reasoning, as well as performance on lo…
Haoran Wei, Yaofeng Sun, Yukun Li
We present DeepSeek-OCR 2 to investigate the feasibility of a novel encoder—DeepEncoder V2—capable of dynamically reordering visual tokens upon image semantics. Conventional vision-language models (VLMs) invariably process visual tokens in a rigid raster-scan order (top-left to bottom-right) with fixed positional enco…
Xin Cheng, Xingkai Yu, Chenze Shao, Jiashi Li, Yunfan Xiong Yi Qian, Jiaqi Zhu, Shirong Ma, Xiaokang Zhang, Jiasheng Ye, Qinyu Chen, Chengqi Deng, Jiping Yu, Damai Dai, Zhengyan Zhang, Yixuan Wei, Yixuan Tan, Wenkai Yang, Runxin Xu, Yu Wu, Zhean Xu, Xuanyu Wang, Muyang Chen, Rui Tian, Xiao Bi, Zhewen Hao, Shaoyuan Chen, Huanqi Cao, Wentao Zhang, Anyi Xu, Huishuai Zhang, Dongyan Zhao, Wenfeng Liang
Speculative decoding accelerates Large Language Model (LLM) inference by decoupling draft generation from target verification. While recent parallel drafters efficiently propose long token sequences in a single forward pass, they suffer from rapid acceptance decay due to a lack of inter-token dependencies. Furthermore…
Lorenzo Mur-Labadia, Matthew Muckley, Amir Bar et al.
We present **V-JEPA 2.1**, a family of self-supervised models that learns dense, high-quality representations for visual scenes in both images and videos, while retaining strong global scene understanding. V-JEPA 2.1 combines four key ingredients: (i) a **Dense Predictive Loss**, a masking-based objective in which all…
DiffusionGemma Team, Google DeepMind
We introduce DiffusionGemma, an experimental open-weight language model that uses discrete diffusion to generate text at exceptionally high speed. Rather than decoding one token at a time, DiffusionGemma iteratively refines blocks of 256 tokens in parallel, avoiding the sequential decoding bottleneck of conventional a…
Madhuri Shanbhogue, Zhe Li, Shanfeng Zhang et al.
We introduce Gemini Embedding 2, a native multimodal embedding model that allows embedding video, audio, image, and text modalities in a unified representation space. We leverage the multimodal capabilities of Gemini to produce embeddings for arbitrary combinations of interleaved inputs across all these modalities tha…
Peng Xia, Jianwen Chen, Hanyang Wang et al.
Large Language Model (LLM) agents have shown stunning results in complex tasks, yet they often operate in isolation, failing to learn from past experiences. Existing memory-based methods primarily store raw trajectories, which are often redundant and noise-heavy. This prevents agents from extracting high-level, reusab…
Gemma Team, Google DeepMind\authfootnotemark1
We introduce Gemma 4, a new generation of open-weight, natively multimodal language models in the Gemma model family. Designed to advance compute efficiency and reasoning, the Gemma 4 model suite features dense and Mixture-of-Experts architectures, ranging from 2.3B to 31B parameters. Alongside improved vision and aud…
Zhixia Zhang, Zixuan Huang, Gongxun Li et al.
We introduce **H**eterogeneous **A**gent **C**ollaborative **R**einforcement **L**earning (**HACRL**), a new Reinforcement Learning from Verifiable Reward (RLVR) problem that addresses the inefficiencies of isolated multi-agent on-policy optimization. HACRL enables collaborative optimization with independent execution…
Jenny Zhang, Bingchen Zhao, Wannan Yang et al.
Self-improving AI systems aim to reduce reliance on human engineering by learning to improve their own learning and problem-solving processes. Existing approaches to recursive self-improvement typically rely on fixed, handcrafted meta-level mechanisms, which fundamentally limit how fast such systems can improve. The D…
Yinjie Wang, Xuyang Chen, Xiaolong Jin et al.
Every agent interaction generates a **next-state signal**, namely the user reply, tool output, terminal or GUI state change that follows each action, yet no existing agentic RL system recovers it as a live, online learning source. We present **OpenClaw-RL**, a framework that employs next-state signals to optimize pers…
Zhengxi Lu, Zhiyuan Yao, Jinyang Wu, Chengcheng Han, Qi Gu
Agent skills, structured packages of procedural knowledge and executable resources that agents dynamically load at inference time, have become a reliable mechanism for augmenting LLM agents. Yet inference-time skill augmentation is fundamentally limited: retrieval noise introduces irrelevant guidance, injected skill c…
Yaorui Shi, Yuxin Chen, Zhengxi Lu et al.
A persistent skill library allows language model agents to reuse successful strategies across tasks. Maintaining such a library requires three coupled capabilities. The agent selects a relevant skill, utilizes it during execution, and distills new skills from experience. Existing methods optimize these capabilities in…
Shuo Yang , Jinyang Wu , Zhengxi Lu, Yuhao Shen, Fan Zhang, Lang Feng,
Outcome-based reinforcement learning provides a stable optimization backbone for language agents, but its sparse trajectory-level rewards provide little guidance on which intermediate decisions should be reinforced or suppressed. On-policy self-distillation offers dense token-level supervision, yet existing skill-cond…
Zhengxi Lu, Zhiyuan Yao, Zhuowen Han, Zi-Han Wang, Jinyang Wu
Reinforcement learning (RL) has emerged as a central paradigm for post-training LLM agents, yet its trajectory-level reward signal provides only coarse supervision for long-horizon interaction. On-Policy Self-Distillation (OPSD) complements RL by introducing dense token-level guidance from a teacher branch augmented w…
Zhenyu Hou, Yujiang Li, Jie Tang et al.
Reinforcement learning (RL) is becoming increasingly important for post-training large language models (LLMs). Previous RL pipelines for LLMs were mostly synchronous and batch-interleaved, which is inefficient for long-horizon agentic tasks. Recently, asynchronous RL has emerged as a more efficient alternative by upda…
Jinyang Wu , Shuo Yang , Zhengxi Lu, Fan Zhang, Yuhao Shen, Lang Feng,
Large language models are increasingly trained as interactive agents for long-horizon tasks involving multi-turn interaction, tool use, and environment feedback. Outcome-based reinforcement learning (RL) provides a practical optimization paradigm, but its sparse trajectory-level rewards offer limited guidance on inter…
Chengsong Huang, Zifeng Wang, Rujun Han et al.
LLM agents learn by interacting with environments, yet these environments are hand-built and static: blind to an agent’s weaknesses, and quickly left behind as it improves. While recent environment generation methods attempt to address this, they require domain-specific pipelines, rely on expensive or unreliable verif…
Wei Zhou, Jun Zhou, Haoyu Wang et al.
Data preparation aims to denoise raw datasets, uncover cross-dataset relationships, and extract valuable insights from them, which is essential for a wide range of data-centric applications. Driven by (i) rising demands for application-ready data (e.g., for analytics, visualization, decision-making), (ii) increasingly…
Jingqi Tong, Mingzhe Li, Hangcheng Li, Yongzhuo Yang, Yurong Mou,
Scientific discovery depends on expert judgement and foresight, which we call scientific taste: the ability to judge and propose research ideas with the potential for long-term scientific impact. Scientific taste is largely concentrated among highly experienced researchers, whose expertise is usually limited to a few…
Ruiyan Han, Zhen Fang, Xinyu Sun et al.
While Unified Multimodal Models (UMMs) have achieved remarkable success in cross-modal comprehension, a significant gap persists in their ability to leverage such internal knowledge for high-quality generation. We formalize this discrepancy as Conduction Aphasia, a phenomenon where models accurately interpret multimod…