Proposes a linear arithmetic model with CEGAR-inspired SMT solver for optimal sequential 3D object packing and scheduling.

Pavel Surynek, Vojtěch Bubník, Lukáš Matěna et al.

START enhances reasoning by integrating tools, achieving 63.6% accuracy on GPQA.

Chengpeng Li, Mingfeng Xue, Zhenru Zhang et al.

HybridNorm combines Pre-Norm and Post-Norm to enhance Transformer training stability and performance.

Zhijian Zhuo, Yutao Zeng, Ya Wang et al.

GAGrasp employs geometric algebra with diffusion and physics-based refinement to enhance SE(3) equivariance, improving dexterous grasp generation robustness and efficiency.

Tao Zhong, Christine Allen-Blanchette

ADiT unifies molecule and material generation via a shared latent diffusion framework, achieving SOTA results with high speed.

Chaitanya K. Joshi, Xiang Fu, Yi-Lun Liao et al.

CHOP: Enhancing mobile assistants with constrained high-frequency optimized subtask planning.

Yuqi Zhou, Shuai Wang, Sunhao Dai et al.

Proposes dual-objective optimization (DOOR/W-DOOR) to improve LLM safety alignment, significantly reducing jailbreak success rates while maintaining utility.

Xuandong Zhao, Will Cai, Tianneng Shi et al.

LION-FS enhances real-time video assistant accuracy via Fast & Slow path strategy.

Wei Li, Bing Hu, Rui Shao et al.

Using Sparse Autoencoders (SAE) to extract interpretable features from Gemma-2-2b residual streams enhances AI text detection and interpretability.

Kristian Kuznetsov, Laida Kushnareva, Polina Druzhinina et al.

This study evaluates transformer models for molecular property prediction, showing domain adaptation with physicochemical tasks significantly outperforms larger data scaling.

Afnan Sultan, Max Rausch-Dupont, Shahrukh Khan et al.

CodeIF-Bench evaluates LLM instruction-following in multi-turn code generation, focusing on context management and nine verifiable instruction types.

Peiding Wang, Li Zhang, Fang Liu et al.

Spike-based alignment learning (SAL) uses spike timing statistics to correct synaptic asymmetry, solving the weight transport problem in neural networks.

Timo Gierlich, Andreas Baumbach, Akos F. Kungl et al.

The paper surveys personalized generation, proposing a multi-level taxonomy and exploring technical advances and applications.

Yiyan Xu, Jinghao Zhang, Alireza Salemi et al.

CMMLoc employs Cauchy Mixture Models within a Transformer framework to model partial relevance, achieving state-of-the-art localization accuracy on KITTI360Pose with 83% within 5m.

Yanlong Xu, Haoxuan Qu, Jun Liu et al.

The GA-Rollback framework employs generator-assistant collaboration with stepwise rollback to significantly improve LLM agent reasoning accuracy, achieving up to 10.5× success rate improvements on benchmark tasks.

Xingzuo Li, Kehai Chen, Yunfei Long et al.

EpicPRM framework constructs high-quality math reasoning dataset Epic50k using adaptive binary search, enhancing PRM performance.

Wei Sun, Qianlong Du, Fuwei Cui et al.

PD-VLA employs parallel fixed-point iteration to accelerate action chunking in VLA models, achieving 2.52× speedup while maintaining performance.

Wenxuan Song, Jiayi Chen, Pengxiang Ding et al.

Proposes a PFC-HPC interaction-based reinforcement learning model for goal-directed memory control and generalization.

Yicong Zheng, Nora Wolf, Charan Ranganath et al.

This paper introduces ToolRet, a benchmark for large language model tool retrieval, revealing poor IR model performance and proposing 200k training samples for improvement.

Zhengliang Shi, Yuhan Wang, Lingyong Yan et al.

CoPL enhances LLM personalization using graph-based collaborative filtering and LoRA experts, improving preference estimation.

Youngbin Choi, Seunghyuk Cho, Minjong Lee et al.