Daily AI Papers — August 05, 2026

1. MerchantBench: Benchmarking LLM Agents for Long-Term Coherence in E-Commerce Operations

Authors: Qiming Shi, Yulong Tao, Linbo Jin, Zhaolu Kang, Yibo Dou, Jiawen Zhu, Tianjun Pan, Shaokang Fu, Chengyu Wang, Siyue Li, Yaping Cheng, Di Weng, Chengfu Huo arXiv: arxiv.org/abs/2607.28956 Summary: Large language model agents are increasingly evaluated as autonomous tool users, yet most benchmarks focus on bounded tasks with immediate success criteria. Real-world deployments often require Long-Term Coherence, the capacity to preserve purposeful behavior across extended horizons while adapting decisions to accumulated evidence. Trending because: 85 HuggingFace upvotes today.

2. JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion

Authors: Yicheng Xiao, Wenxun Dai, Xinran Qin, Lin Song, Maoquan Zhang, Hang Xu, Yukang Chen, Yitong Li, Guohui Zhang, Yuan Zhang, Xuying Zhang, Tommy Zhang, Jianlong Yuan, Peihao Li, Shuai Lu, Siming Fu, Chuyang Zhao, Xin Han, Jie Huang, Wenbo Li, Guoqing Ma, Wei Huang, Xiaojuan Qi, Haoyang Huang, Nan Duan arXiv: arxiv.org/abs/2608.03974 Summary: Real-time video editing requires low-latency causal generation with bounded computational resources while preserving source fidelity and long-term temporal consistency. We present JoyAI-Video-Edit, a 16B-parameter autoregressive diffusion framework for real-time, open-ended video editing without access to future frames or a predefined video duration. Trending because: 77 HuggingFace upvotes today.

3. AURORA-LM: Autoencoding Unified Representation for Continuous-Latent Diffusion Language Modeling

Authors: Jiajun Liang, Yucheng Liao, Yukang Cao, Jiazhe Wei, Ken Li, Wende Tan, Jiankun Zhang, ZY Cui, Jingkang Yang, Liucheng Guo, Shiqi Yang, B. Yang, Caifeng Shan, Ziwei Liu, Chenyang Si arXiv: arxiv.org/abs/2608.02602 Summary: Language remains an outlier in generative modeling: while images, video, and audio are increasingly modeled in continuous latent spaces, text generation still relies predominantly on discrete tokens. Existing continuous language models either inherit embedding spaces not designed for joint generation and decoding, or compress autoencoded latents to ease diffusion, sacrificing token-level fidelity. Trending because: 73 HuggingFace upvotes today.

4. Hunyuan3D-Buffalo 1.0: A Unified Multimodal Model for Scalable 3D Generation, Understanding, and Editing

Authors: Junliang Ye, Kenkun Liu, Guocun Wang, Yang Li, Yansong Qu, Chunshi Wang, Jingwei Xu, Yunhan Yang, Zibo Zhao, Jiachen Xu, Jiaao Yu, Lifu Wang, Zhihao Liang, Zhuo Chen, Chunchao Guo arXiv: arxiv.org/abs/2608.02711 Summary: Recent advances in image generation have demonstrated the potential of unified multimodal models that integrate understanding, generation, and editing. However, unified 3D modeling remains constrained by scarce multimodal data, particularly the lack of large-scale and geometrically consistent editing data. Trending because: 71 HuggingFace upvotes today.

5. InfiniSplat: Implicit Gaussian Decoding for Large-Baseline Monocular View Synthesis

Authors: Jiawei Wang, Hao Yu, Yongzhen Hu, Xinyi Yang, Tao Ni, Xin Zhan, Junbo Chen, Xiaowei Zhou, Ruizhen Hu, Sida Peng arXiv: arxiv.org/abs/2608.02437 Summary: Single-image feed-forward 3D Gaussian Splatting (3DGS) aims to directly generate a renderable 3D scene representation from one input image, avoiding the cost of multi-view capture and per-scene optimization. However, existing methods are often constrained by a pixel-aligned representation, where Gaussians are predicted from fixed image-grid locations. Trending because: 56 HuggingFace upvotes today.

6. Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent

Authors: Zhen Fang, Yu Zeng, Wenxuan Huang, Yiming Zhao, Shiting Huang, Tianfei Ren, Qi Lu, Qingnan Ren, Qisheng Su, Lionel Z. Wang, Qingyu Yin, Shuang Chen, Zehui Chen, Lin Chen, Zhenfei Yin, Yao Hu, Shaohui Lin, Wanli Ouyang, Shaosheng Cao, Feng Zhao arXiv: arxiv.org/abs/2608.03979 Summary: We introduce Video-DeepResearch (Video-DR), extending multimodal agents from static images to continuous video streams, a setting that demands dense spatiotemporal grounding coupled with open-web exploration. Trending because: 45 HuggingFace upvotes today.

7. Knowledge-Geometry Decoupling: Refreshable Pretrained Transfer for Streaming Recommendation

Authors: Zixuan Wang, Yuhong Chen, Yuxuan Zhu, Guidong Lei, Zhiluohan Guo, Yu Zhao, Kun Wang, Bangyang Hong, Kangle Wu, Yabo Ni, Anxiang Zeng, Cong Fu, Hui Li arXiv: arxiv.org/abs/2608.02738 Summary: Industrial recommenders increasingly adopt the pretrain-then-transfer paradigm, yet behavioral distribution drift raises two questions: what to learn from behavior sequences, and how to transfer the learned knowledge while the pretrained model is continually refreshed. To resolve them, we propose Knowledge-Geometry Decoupling (KGD). Trending because: 38 HuggingFace upvotes today.

8. PCSD: Persistent Consistency for Self-Distillation in Agentic Reinforcement Learning

Authors: Chunji Lv, Yangguang Wei, Junlin Liu, Yang Gao, Ming Liu, Xinming Wang, Jinyang Wu, Guoren Wang, Changsheng Li arXiv: arxiv.org/abs/2608.01837 Summary: Large language model agents have shown strong potential in complex interactive tasks, yet their reinforcement learning (RL) is often hindered by sparse rewards, as a long multi-turn trajectory may receive only a single outcome-level signal. On-policy self-distillation (OPSD) provides dense token-level supervision from a privileged teacher, but the teacher may not be reliable at every position. Trending because: 37 HuggingFace upvotes today.

9. Quo Vadis, World Modeling?

Authors: Yu Yang, Xuemeng Yang, Licheng Wen, Lingdong Kong, Xiaobin Hu, Dongyue Lu, Wei Chow, Xiyan Huang, Yuxiang Feng, Yue Liao, Jianbiao Mei, Daocheng Fu, Rong Wu, Pinlong Cai, Ran Yi, Ying Tai, Jiangning Zhang, Botian Shi, Yong Liu, Shuicheng Yan arXiv: arxiv.org/abs/2608.02713 Summary: Continually improving agents require dynamic interaction feedback beyond static supervision, yet direct real-environment interaction is costly, slow, unsafe, and hard to parallelize. World modeling offers a natural intermediate proxy that allows agents to query lower-cost, more controllable feedback before committing to real actions. Trending because: 29 HuggingFace upvotes today.

10. PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents

Authors: Shuhan Xue, Zixin Ding, Yichen Shen, Yinjie Wang, Zhenfei Yin, Yingcheng Wu, Yuxin Chen, Mengdi Wang, Ling Yang arXiv: arxiv.org/abs/2608.04003 Summary: Recursive self-improvement requires agents to turn accumulated experience into better future behavior. Personal AI agents offer a concrete setting for studying this capability because they retain preferences, task histories, tool routines, and learned skills across sessions. Trending because: 28 HuggingFace upvotes today.

11. OmniPack: Unified Token Compression for Efficient Omni-modal Large Language Models

Authors: Wanshun Su, Yang Shi, Feihu Liu, Ziwen Yu, Yan Min, Zhuoran Zhang, Qixun Wang, Haotian Wang, Shixuan Liu, Yuanxing Zhang, Peng Wu, Chengfu Huo, Liang Ding arXiv: arxiv.org/abs/2608.03812 Summary: Omni-modal large language models (Omni-LLMs) have achieved remarkable performance on audio-visual understanding tasks, but processing long and highly redundant visual and audio token sequences incurs substantial computational overhead, demanding aggressive token compression for efficient deployment. Trending because: 24 HuggingFace upvotes today.

12. LLaDA MoE v2: Scaling Mixture-of-Experts Diffusion Language Models

Authors: Fengqi Zhu, Shaoxuan Xu, Jingyang Ou, Zebin You, Yipeng Xing, Huabin Liu, Xiaolu Zhang, Jun Zhou, Zhenzhong Lan, Yankai Lin, Wayne Xin Zhao, Jianguo Li, Chongxuan Li, Ji-Rong Wen arXiv: arxiv.org/abs/2608.03457 Summary: Diffusion language models (dLLMs) offer an alternative to autoregressive (AR) language modeling, yet the scaling behavior of Mixture-of-Experts (MoE) dLLMs remains poorly understood. We systematically characterize how optimization hyperparameters, compute allocation, and architecture scale for MoE dLLMs, identifying quantitative differences from scaling trends previously reported for AR models. Trending because: 24 HuggingFace upvotes today.

13. Any-OPD: Heterogeneous On-Policy Distillation for Flow-Matching Models via Representation-Space Bridging

Authors: Siming Fu, Zheming Fu, Ruizhe He, Hualiang Wang, Jie Huang, Xiaoxiao Ma, Mingchen Zhong, Weihu Huang, Xiaoxuan He, Haojun Xu arXiv: arxiv.org/abs/2608.03316 Summary: On-policy distillation, in which a teacher corrects samples that the student itself generates, presupposes that the two models speak the same language: identical VAE latents, matching architectures, and a common timestep grid. Trending because: 23 HuggingFace upvotes today.

14. CAPEval: A Decoupled Caption Evaluation across Understanding and Generation

Authors: Zhipeng Liu, Haochen Wang, Zhaoxiang Zhang arXiv: arxiv.org/abs/2608.02589 Summary: Captions serve as a primary supervision signal for both multimodal understanding and text-to-image generation. However, previous evaluations treat the caption quality as a single scalar objective, which conflates two distinct properties: (1) how much visual information a caption covers and (2) how reliably the image supports its stated claims. Trending because: 21 HuggingFace upvotes today.

15. MemSFT: Mitigating Alignment Tax with an External Parametric Memory

Authors: Jiarui Wang, Xiang Shi, Jiaqi Cao, Rubin Wei, Xiquan Wang, Hao Sun, Jingzhi Wang, Zhiqi Yang, Qipeng Guo, Bowen Zhou, Zhouhan Lin arXiv: arxiv.org/abs/2607.25614 Summary: Adapting Large Language Models (LLMs) to specialized domains often incurs an alignment tax, as fine-tuning on domain-specific tasks can cause catastrophic forgetting and substantially degrade performance on general tasks. We propose MemSFT, which mitigates the alignment tax by decoupling domain specialization from backbone parameter updates through a plug-and-play parametric memory. Trending because: 21 HuggingFace upvotes today.

16. SkillJack: Persistent Skill Backdoors in Self-Evolving Agents

Authors: Zonghao Ying, Xiangfan Wu, Huiyu Wu, Xing Zheng, Huangsheng Cheng, Xiaorong Shi, Jing Guo arXiv: arxiv.org/abs/2608.03509 Summary: Self-evolving agents increasingly convert interaction histories into reusable skills that persist beyond individual tasks. While prior work studies memory and retrieval poisoning, such attacks only affect agents when poisoned records are retrieved as context. Trending because: 19 HuggingFace upvotes today.

17. UniWorld-Design: From Pixel Generation to Layer-Native Design

Authors: Zongjian Li, Zhiyuan Yan, Chenxu Bai, Chen Chen, Haoxiang Sun, Shaodong Wang, Feize Wu, Shenghai Yuan, Bin Lin, Zheyuan Liu, Yuwei Niu, Li Yuan arXiv: arxiv.org/abs/2608.03971 Summary: We introduce UniWorld-Design, a framework that redefines image generation from flat pixel synthesis to structured visual composition, with semantic RGBA layers as the atomic units of generation, understanding, and editing. Our key insight is that pixels define how an image is rendered, whereas layers define how an image is created, understood, and edited. Trending because: 19 HuggingFace upvotes today.

18. To Add Is Machine, To Delete Is Human: Measuring and Mitigating Deletion Avoidance in LLM Code Editing

Authors: Amir M. Ebrahimi, Mohammed Mehedi Hasan, Aaditya Bhatia, Gopi Krishnan Rajbahadur, Ahmed E. Hassan arXiv: arxiv.org/abs/2607.28887 Summary: Large language models increasingly write and repair production code, yet evidence is mounting that their test-passing patches leave codebases harder to maintain. We identify one concrete source: deletion avoidance, the systematic tendency to retain code that an intended edit requires removing. Trending because: 19 HuggingFace upvotes today.

19. TurnSight: Turn-Level Hindsight Self-Distillation for Tool-Integrated Reasoning

Authors: Changle Qu, Sunhao Dai, Hengyi Cai, Yuqi Zhou, Xinran Chen, Simon, Jun Xu arXiv: arxiv.org/abs/2608.04007 Summary: Tool-Integrated Reasoning (TIR) enables LLMs to solve complex tasks through iterative tool interactions. However, existing reinforcement learning methods often rely on trajectory-level supervision, limiting fine-grained credit assignment in long-horizon TIR scenarios. Trending because: 17 HuggingFace upvotes today.

20. GROVE: Growing and Reasoning over Temporally Stratified Memory from Streaming Video Experience

Authors: Sitong Gong, Caixin Kang, Tianyu Yan, Guo Chen, Bo Zheng, Kaipeng Zhang, Yunzhi Zhuge, Xiang Ruan, Huchuan Lu, Yifei Huang arXiv: arxiv.org/abs/2608.02392 Summary: A wearable assistant should both answer questions about its visual history and recognize when that history is useful to the present situation. Existing video-memory systems primarily support question-conditioned recall, whereas proactive assistants typically use separate memory and control mechanisms. Trending because: 14 HuggingFace upvotes today.