FT-Dojo automates LLM fine-tuning with FT-Agent, excelling in 10 out of 13 tasks.

Qizheng Li, Yifei Zhang, Xiao Yang et al.

Gome employs gradient-based optimization for ML engineering tasks, achieving 35.1% medal rate on MLE-Bench, surpassing tree search methods.

Yifei Zhang, Xu Yang, Xiao Yang et al.

LinkVLA improves autonomous driving instruction accuracy to 91.01 by unifying language and action tokens.

Xinyang Wang, Qian Liu, Wenjie Ding et al.

UniTalking framework uses Multi-Modal Transformer Blocks to generate high-fidelity speech and synchronized video, surpassing existing open-source methods.

Hebeizi Li, Zihao Liang, Benyuan Sun et al.

VACO employs advantage alignment and TV filtering to mitigate policy lag in asynchronous RL, improving robustness and convergence.

Homayoun Honari, Roger Creus Castanyer, Michael Przystupa et al.

Introduces RMBench benchmark and Mem-0 policy, systematically evaluating memory capabilities in robotic manipulation with success rates up to 42%.

Tianxing Chen, Yuran Wang, Mingleyang Li et al.

ReGFT leverages human reference solutions to synthesize positive trajectories, significantly alleviating reward sparsity and boosting RL-based mathematical reasoning performance.

Yangzhen Wu, Shanda Li, Zixin Wen et al.

Semantic XPath uses tree-structured memory for conversational AI, improving performance by 176.7% with only 9.1% of tokens.

Yifan Simon Liu, Ruifan Wu, Liam Gallagher et al.

vEcho leverages large language models with memory and reasoning to shift from passive vulnerability verification to proactive discovery, achieving 65% detection rate and reducing false positives to 59.78%.

Mingcheng Jiang, Jiancheng Huang, Jiangfei Wang et al.

Kernel-based LMI approach for solving nonlinear HJB equations, integrating Riccati-Hessian equality for accurate approximation.

Boumediene Hamzi, Umesh Vaidya

FlexiMMT achieves multi-object multi-motion transfer using Motion Decoupled Mask Attention, improving motion fidelity by 23%.

Yuze Li, Dong Gong, Xiao Cao et al.

SOLACE enhances text-to-image generation quality using intrinsic self-confidence signals, reducing external supervision needs.

Seungwook Kim, Minsu Cho

Minimalist Compliance Control uses motor current signals for sensorless compliance control, applicable to various robots.

Haochen Shi, Songbo Hu, Yifan Hou et al.

RLAR uses LLMs to autonomously retrieve and synthesize reward models, improving multi-task RL performance by 10-60%.

Andrew Zhuoer Feng, Cunxiang Wang, Bosi Wen et al.

K²-Agent enhances mobile device control via co-evolving 'know-what' and 'know-how', achieving a 76.1% success rate.

Zhe Wu, Donglin Mo, Hongjin Lu et al.

Proposes Riemannian heat flow-based hypergraph neural network with adaptive local exchanger, addressing long-range dependencies in heterophilic and homophilic hypergraphs.

Li Sun, Ming Zhang, Wenxin Jin et al.

SWE-Hub unifies environment automation, scalable synthesis, and diverse task generation to support continuous, executable software engineering tasks.

Yucheng Zeng, Shupeng Li, Daxiang Dong et al.

Proposes IFCodeEvolve, a co-evolution framework using actor-schema interaction and MCTS to generate instruction-code data, boosting model performance.

Tinglin Huang, Bo Chen, Xiao Zhang et al.

AxProverBase achieves competitive performance with iterative proof refinement and context management in a simplified architecture.

Borja Requena, Austin Letson, Krystian Nowakowski et al.

Introduced smooth-rank algorithm to optimize bipartite ranking with continuous distributions, enhancing ROC curve precision.

James Cheshire, Stephan Clémençon