LLMs combined pairwise comparison and fine-tuning improve scalar construct measurement in social science, surpassing direct scoring.

Hauke Licht, Rupak Sarkar, Patrick Y. Wu et al.

Introduced PORT, a training-free online routing algorithm, achieving 3.55x performance and 1.85x cost efficiency improvements.

Fangzhou Wu, Sandeep Silwal

DARLING framework integrates a semantic classifier into reinforcement learning to jointly optimize language model diversity and quality, significantly improving creative and mathematical tasks.

Tianjian Li, Yiming Zhang, Ping Yu et al.

SimpleTIR algorithm boosts AIME24 score from 22.1 to 50.5 by filtering void turns.

Zhenghai Xue, Longtao Zheng, Qian Liu et al.

U-Arm, a low-cost ($50-$56) universal teleoperation platform, uses mechanical and control optimizations to enable efficient data collection across most commercial robots.

Yanwen Zou, Zhaoye Zhou, Chenyang Shi et al.

RDIT combines point estimation and residual diffusion, optimizing CRPS, outperforming baselines on 8 datasets with faster inference.

Chih-Yu Lai, Yu-Chien Ning, Duane S. Boning

MME-SID combines multimodal quantization and semantic-ID initialization; on Amazon Beauty, tau rises from 0.0550 to 0.3714.

Yuhao Wang, Junwei Pan, Xinhang Li et al.

Proposed SoTPDEG framework for multi-graph data, preserving high-frequency info with spectral decomposition and stability guarantees.

Aref Einizade, Fragkiskos D. Malliaros, Jhony H. Giraldo

Attention-based encoder-decoder deep RL (AEDM) optimizes drone routes, achieving 20-71% better solutions in seconds for post-disaster road assessment.

Huatian Gong, Jiuh-Biing Sheu, Zheng Wang et al.

Kwai Keye-VL 1.5 employs Slow-Fast video encoding, extends context to 128K tokens, combined with multi-stage pretraining and post-training, greatly enhancing video understanding.

Biao Yang, Bin Wen, Boyang Ding et al.

Unified geometric framework for nonlinear RL, integrating reward, safety, and diversity via occupancy measure optimization.

Nikola Milosevic, Nico Scherf

FantasyHSI enables 4D human synthesis in any scene using a graph-based multi-agent framework, significantly improving task completion.

Lingzhou Mu, Qiang Wang, Fan Jiang et al.

POINTS-Reader employs a two-stage, distillation-free framework using synthetic data and self-improvement, achieving state-of-the-art document conversion performance.

Yuan Liu, Zhongyin Zhao, Le Tian et al.

Proposed BiLSTM-AM-VMD framework integrates multimodal data for early liver cancer diagnosis, achieving AUC of 0.963, outperforming baseline models.

Cheng Cheng, Zeping Chen, Xavier Wang

Multi-modal ML framework combining MRI and biomarkers achieves 78.2% C-index for early brain tumor recurrence prediction.

Cheng Cheng, Zeping Chen, Rui Xie et al.

GPSToken employs Gaussian parameterization for spatially adaptive image tokenization, achieving state-of-the-art FID=1.50 in image generation with 128 tokens.

Zhengqiang Zhang, Rongyuan Wu, Lingchen Sun et al.

OmniDPO reduces omni-modal hallucination using a preference optimization framework, enhancing multimodal reasoning.

Junzhe Chen, Tianshu Zhang, Shiyu Huang et al.

Using kernel methods for nonlinear augmentation in dimensionality reduction, improving accuracy and reducing training costs.

Alejandro N. Diaz, Jacob T. Needels, Irina K. Tezaur et al.

Defines and quantifies algorithm adaptation bias in online recommender system experiments, validated through real-world A/B tests, emphasizing ecosystem feedback effects.

Chen Zheng, Zhenyu Zhao

Safe-LLaVA reduces privacy leakage by removing biometric info from LLaVA dataset.

Younggun Kim, Sirnam Swetha, Fazil Kagdi et al.