Proposed an LLM-based text animation editing system with inline suggestions and chat interaction, reducing non-professional users' editing time by 30%.

Bao Zhang, Zihan Li, Zhenglei Liu et al.

ClaimSpect uses retrieval-augmented hierarchical analysis to decompose nuanced claims, identifying multi-angle evidence and perspectives.

Priyanka Kargupta, Runchu Tian, Jiawei Han

Study reveals complexities in LLMs' character-level spelling tasks, showing reliance on mid-to-high Transformer layers for reconstructing character information.

Tatsuya Hiraoka, Kentaro Inui

DoublyAware decomposes planning and policy uncertainties, improving sample efficiency and robustness in humanoid locomotion via conformal prediction and group-relative policy constraints.

Khang Nguyen, An T. Le, Jan Peters et al.

StepProof enhances verification success rates and efficiency through step-by-step validation of natural language mathematical proofs.

Xiaolin Hu, Qinghua Zhou, Bogdan Grechuk et al.

DiMo-GUI combines modality decoupling and dynamic zooming to improve GUI grounding accuracy without extra training, boosting performance over baseline models.

Hang Wu, Hongkai Chen, Yujun Cai et al.

Rectified flow turns 3D hand trajectories into interaction sounds that are often indistinguishable from real audio.

Yiming Dou, Wonseok Oh, Yuqing Luo et al.

Uni-DPO enhances LLM performance by dynamically reweighting samples, outperforming Claude 3 Opus by 6.7 points on Arena-Hard.

Shangpin Peng, Weinong Wang, Zhuotao Tian et al.

Proposes dual volume packing for efficient part-level 3D generation from a single image, avoiding segmentation priors.

Jiaxiang Tang, Ruijie Lu, Zhaoshuo Li et al.

PersonaLens uses multi-task, multi-domain user profiles and two LLM agents to evaluate personalization in task-oriented dialogue systems.

Zheng Zhao, Clara Vania, Subhradeep Kayal et al.

Proposes R2SE framework combining hard-case identification and reinforcement fine-tuning to enhance end-to-end autonomous driving.

Haochen Liu, Tianyu Li, Haohan Yang et al.

Identifies three common pitfalls in KL gradient estimation for RL, emphasizing the correct full-sequence approach.

Yunhao Tang, Rémi Munos

Introduces Subspace-Constrained Randomized Coordinate Descent (SC-RCD) leveraging Nyström low-rank approximation, robust against spectral outliers, for large-scale positive semidefinite linear systems.

Jackie Lok, Elizaveta Rebrova

LPO combines information entropy and dynamic distance rewards to enhance GUI interaction accuracy, achieving state-of-the-art results.

Jiaqi Tang, Yu Xia, Yi-Feng Wu et al.

e3 trains models with chain-based exploration, enabling extrapolation to twice the training inference length, significantly improving reasoning performance beyond training limits.

Amrith Setlur, Matthew Y. R. Yang, Charlie Snell et al.

MIL models show superior transfer across organs, boosting performance by 9.8%.

Daniel Shao, Richard J. Chen, Andrew H. Song et al.

Proposes a trust-based resilient multi-robot protocol with dynamic λt to ensure robustness against malicious robots, validated through theoretical proofs and experiments.

Luca Ballotta, Áron Vékássy, Stephanie Gil et al.

Comparative study of auto-regressive decoding and conditional flow matching for text-to-music generation.

Or Tal, Felix Kreuk, Yossi Adi

Proposes 4D Gaussian Transformer (4DGT) for real-time dynamic scene reconstruction from monocular videos, significantly improving speed and accuracy.

Zhen Xu, Zhengqin Li, Zhao Dong et al.

Self Forcing introduces autoregressive training with distribution matching, enabling real-time video generation with high quality and low latency, outperforming traditional slow diffusion models.

Xun Huang, Zhengqi Li, Guande He et al.