Paper Archive

[object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

Unified multimodal models (UMMs) aim to handle perception and generation in a single model. Yet existing UMMs still rely on a frozen, separately pretrained VAE for image generation, imposing a structural bottleneck. Naively removing it introduces a quality gap, as the model must learn both high-leve...

[object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

Connector-based video unified models have demonstrated strong capability in instruction-grounded video synthesis, but integrating a large high-fidelity generator into the unified training loop is computationally prohibitive, limiting achievable visual quality. We therefore propose Lumos-Nexus, a tra...

[object Object], [object Object], [object Object] 5/29/2026 huggingface

Video vision-language models (VLMs) are increasingly used in long-horizon and streaming settings, yet most video encoders still rely on spatiotemporal self-attention, causing compute and latency to grow quadratically with the number of frames. Existing efficiency methods improve scalability but ofte...

[object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

Long-context reasoning remains a central challenge for large language models, which often fail to locate and integrate key information in extensive distracting content. Reinforcement learning with verifiable rewards (RLVR) has shown promise for this task, yet existing methods are limited by low-conf...

[object Object], [object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

GPU kernels are the workhorse of modern deep learning, and optimizing them (via evolutionary search or coding agents) usually requires repeated measurement on target hardware. While these measurements provide the ground-truth signal necessary for kernel search, they are costly, because each evaluati...

[object Object], [object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

Large language models are increasingly deployed in multi-turn interactive settings where users or environments can iteratively provide lightweight feedback. Unfortunately, optimizing such behavior presents a sharp dilemma in practice: online reinforcement learning is able to effectively address mult...

[object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

Self-play can train language models without external supervision. However, existing methods require rule-checkable answers, leaving open-ended tasks dependent on curated prompts or frontier-model judges. We introduce SCOPE, a data-free self-play framework for open-ended tasks that co-evolves two pol...

[object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

Statutory references are central to legal language understanding, but are difficult to process automatically, as they appear in compact and variable surface forms, may combine multiple targets, use special abbreviations, and often point to lower-level units. Existing tools for German focus either on...

[object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

Recent advances in video generative models have promoted rapid progress in controllable world models. However, maintaining fine-grained spatio-temporal consistency under long-horizon reasoning remains a key challenge. In this work, we move beyond explicit 3D memory and coarse frame-level implicit mo...

[object Object], [object Object], [object Object], [object Object], [object Object] 5/29/2026 huggingface

LLM agents are increasingly expected not only to complete isolated tasks, but also to carry bounded representations of human expertise, judgment, and interaction style. Building such person-grounded agents remains difficult because actionable knowledge associated with a person or role is usually emb...