TAAM leverages lightweight Neural Synapse Modulators and Anchored Multi-hop Propagation for replay-free, resource-efficient continual graph learning, outperforming SOTA with zero forgetting.

Jingtao Liu, Xinming Zhang

Analyzed 40,285 skills to study how Claude skills extend large language model functionality.

George Ling, Shanshan Zhong, Richard Huang

OGPSA uses orthogonal gradient projection to mitigate capability loss during LLM safety alignment, improving safety-utility trade-off by over 9%.

Guanglong Sun, Siyuan Zhang, Liyuan Wang et al.

rePIRL learns process reward models via inverse RL to enhance LLM reasoning.

Xian Wu, Kaijie Zhu, Ying Zhang et al.

Proposes Learnable Chernoff Baselines (LCBs) for inference-time model alignment, reducing queries by 7x, with total variation guarantees.

Sunil Madhow, Yuchen Liang, Ness Shroff et al.

ViCA employs vision-only cross-attention, reducing visual computation to 4% while maintaining 98% accuracy, greatly improving efficiency.

Wenjie Liu, Hao Wu, Xin Qiu et al.

Introduced Epistemic Ledger to diagnose and reduce illusory completion in search agents, lowering underverification by up to 26.5%.

Dayoon Ko, Jihyuk Kim, Sohyeon Kim et al.

ZoomDet introduces offset prediction and corner-aligned box transformation for UAV small-object detection, achieving 8.4% mAP gain with minimal latency.

Tao Wang, Chenyu Lin, Chenwei Tang et al.

ShaPO enhances LLM safety alignment robustness via selective geometry control, outperforming benchmarks.

Yonghui Yang, Wenjian Tao, Jilong Liu et al.

RuleFlow employs a three-stage pipeline to convert LLM-discovered code optimizations into reusable rewrite rules, achieving up to 4.3× speedup on Pandas benchmarks.

Avaljot Singh, Dushyant Bharadwaj, Stefanos Baziotis et al.

ANCHOR uses branch-point expansion to generate diverse GUI trajectories, improving model fine-tuning success rates by over 4% on benchmarks.

Jinbiao Wei, Yilun Zhao, Kangqi Ni et al.

Proposes PlugMem, a task-agnostic plugin memory module using knowledge graphs, outperforming task-specific methods with higher information density.

Ke Yang, Zixi Chen, Xuan He et al.

VenusBench-Mobile evaluates mobile GUI agents through user-intent-driven tasks and capability-oriented annotations, revealing their shortcomings in real environments.

Yichen Gong, Zhuohan Cai, Sunhao Dai et al.

Using thermodynamic entropy production, proposes EDS and WDS schedules, significantly improving discrete diffusion sampling efficiency.

Alberto Foresti, Mustapha Bounoua, Giulio Franzese et al.

SNIP, Wanda, SafeNeuron, and NLSR produced IoUs of 0.01–0.72, revealing no stable dataset-agnostic safety region.

Zongmin Li, Jian Su, Farah Benamara et al.

Proposes Soft FB, a maximum entropy variant for zero-shot RL, enabling direct optimization of general differentiable utilities via offline data and low-dimensional search.

Marco Bagatella, Thomas Rupf, Georg Martius et al.

Introduced an Integer Linear Programming approach for geometrically consistent partial-partial shape matching, significantly improving matching accuracy.

Viktoria Ehm, Paul Roetzer, Florian Bernard et al.

TokenMixer-Large introduces mixing-reverting and sparse MoE, scaling models to 15B parameters, achieving significant offline and online performance gains.

Yuchen Jiang, Jie Zhu, Xintian Han et al.

DriveWorld-VLA unifies vision-language-action with world models in latent space, achieving 91.3 PDMS on NAVSIMv1 and 0.16 collision rate on nuScenes.

Feiyang jia, Lin Liu, Ziying Song et al.

AgentCPM-Explore achieves 97.09% GAIA accuracy via parameter merging, reward denoising, and context refinement in a 4B model.

Haotian Chen, Xin Cong, Shengda Fan et al.