The list as it stood that day: the papers researchers were upvoting on Hugging Face's daily papers. Papers stay on arXiv.
When post-training emphasizes one readout, improvements can therefore concentrate on that interface while transfer to others remains weak or even degrades; we call this observed specialization interface lock-in.
Yihao Hu, Yitong Lin, Pan Wang and 8 moreWe present Hy-Nomas, a video world model with an 8B diffusion backbone that supports free-form text instructions and continuous camera control across multiple interaction turns.
Qiaoling Chen, Jiaxiang Cheng, Xueqing Deng and 35 moreWe introduce Rollback, a policy improvement paradigm that turns world models from passive simulators into counterfactual engines: it rolls back to critical decision points, branches into alternative actions, and contrasts their outcomes to attribute success to individual…
Jiayi Luo, Yang Zhang, Liliang Chen and 13 moreWe introduce ThinkWorld, a VLA learning framework that grounds action reasoning in world dynamics and strengthens it through reinforcement learning with world imagination.
Chaehyun Kim, Chi-Pin Huang, Jisoo Kim and 4 moreWe introduce the redistribution sensitivity score (RSS), which identifies large feature responses that recur in a consistent direction, and evaluate the resulting supports on variants excluded from selection.
Ziye Mei, Haotian MaIn this work, we explore Code as Video Caption, a paradigm where executable programs serve as a complementary representation of visual content through graphical primitives, layout rules, and functions of time.
Jiafu Tang, Yifan Xu, Eric Tang and 11 moreThrough a large-scale evaluation across multiple benchmarks and model backbones, we find many dedicated memory systems underperform naive RAG in many cases, indicating a negative net value.
Xuan He, Sining Zhoubian, Ruozhen Yang and 8 moreWe introduce LociArt, a unified paradigm for free-style local video stylization that explicitly specifies where to stylize and what style to apply while preserving non-target content.
Yuxuan Chen, Feng Xie, Jiagao Hu and 4 moreWe introduce Test-Time Memory Evolution (TTME), a plug-and-play test-time adaptation method, as an alternative to TTT.
Yepeng Liu, Shengmai Chen, Xuandong Zhao and 4 moreWe introduce OvisOCR3, a 0.8B model for a broad range of OCR-related tasks. OvisOCR3 unifies document parsing, general OCR, key information extraction (KIE), document figure captioning, and cross-page table reconstruction in a single model.
Yu Xia, Yinglun Li, Yuhui Chen and 12 moreIn long-term interactions with large language model (LLM) agents, users reveal habits through their choices and actions.
Junming Zhang, Rubin Wei, Xu Cheng and 7 moreWe therefore introduce The Trinity, a recursive self-improvement framework coupling world generation with Harness evolution through three roles: Creator, Explorer, and Evolver.
Weidong Tang, Ruiheng Wang, Jingxuan Zhang and 17 moreWe introduce PDRVLA, whose lightweight Physical Dynamics Module extracts physical tokens from vision--language features to condition action generation without online video generation.
Hanqing Wang, Zhaotian Cai, Ruichuan An and 12 moreDiffusion language models (DLMs) promise parallel text generation, but in practice face an unresolved speed-quality trade-off.
Eliezer Shlomi, Eliron Rahimi, Avi Mendelson and 1 moreWe present MatAnyone-E, an efficient extension of MatAnyone 2 that focuses computation where it is most needed.
Shuyun Wang, Shangchen Zhou, Peiqing Yang and 3 moreWe introduce RSI Arena, a two-stage study of autonomous post-training as a step toward recursive self-improvement.
Yue Huang, Yanbo Wang, Lichi Li and 15 moreAn agent that truly knows and acts on behalf of its user has been a long-standing promise of AI assistants.
Yuting Ning, Ruijia Cheng, Sushobhan Ghosh and 7 moreWe show that channel selection needs no calibration data. First, a calibrated threshold does not fix how much a selector reads, because the number of channels passing it varies across inputs; Prox with thresholds calibrated for an 18% read budget actually reads only 12.9% on…
Geunhyoung Kim, Sung Ju HwangWe introduce ReGE, a framework for turning this recorded experience into guidance for a fresh attempt.
Jiarong Liang, Yubo Wang, Zijie Liu and 3 moreWe explore an alternative: general-purpose coding agents that construct images through executable programs.
Hengguang Zhou, Yuanhao Ban, Jiaqi Feng and 3 more