Using influence functions with off-policy estimation and sparse projection, CROPI accelerates large-scale RLVR training by 2.66× with only 10% data per stage.

Erle Zhu, Dazhi Jiang, Yuan Wang et al.

Agent Skills framework is vulnerable to simple prompt injections, leading to data leaks.

David Schmotz, Sahar Abdelnabi, Maksym Andriushchenko

Synthetic data reveals that current MIL models underperform the Bayesian optimal in capturing spatial dependencies, even with 10,000 samples.

Ethan Harvey, Dennis Johan Loevlie, Michael C. Hughes

Combining SHAP and causal discovery algorithms to enhance fault detection accuracy and interpretability in industrial processes.

Pedro Cortes dos Santos, Matheus Becali Rocha, Renato A Krohling

Proposes max@k gradient estimation to optimize Best-of-N sampling, improving code generation performance.

Farid Bagirov, Mikhail Arkhipov, Ksenia Sycheva et al.

SIREN, MFN-Gabor, and MHE compressed thoracic-aorta fields up to 230×, with ~1 mmHg pressure and sub-1.6 mm anatomical errors.

Jubilee Lee, Daniele E. Schiavazzi

ColMAD reframes multi-agent debate as a cooperative game, boosting error detection accuracy by 10% through truthful, informative messaging.

Yongqiang Chen, Gang Niu, James Cheng et al.

SEMPO is a lightweight foundation model for time series forecasting, combining energy-aware spectral decomposition and prompt-based Transformer.

Hui He, Kun Yi, Yuanchi Ma et al.

Proposes Search Self-play (SSP), a self-supervised method that co-evolves question proposing and answering models via multi-turn search and retrieval verification, outperforming baselines.

Hongliang Lu, Yuhang Wen, Pengyu Cheng et al.

COPPOL integrates conformal prediction into semantic segmentation, achieving over 6× hazard coverage and 50% fewer violations in risk-aware RL navigation.

Daniel Bethell, Simos Gerasimou, Radu Calinescu et al.

MAGDrop enhances DNN generalization with momentum-adaptive dropout, achieving 99.52% on MNIST.

Adeel Safder

Extends PAC-Bayes framework to non-compact symmetries, enhancing model generalization.

Armin Beck, Peter Ochs

In-situ Autoguidance uses dropout self-guidance; on ImageNet 512×512 it reaches FID 2.57 with 50k samples.

Enhao Gu, Haolin Hou

Proposes a system-level, multi-modal, multi-agent framework for interpretable and trustworthy time series reasoning, emphasizing structured multi-step inference.

Kanghui Ning, Zijie Pan, Yushan Jiang et al.

SkipV1Former introduces skip connections from first-layer Value heads, reducing KV cache by 25% while improving perplexity.

Zhoutong Wu, Yuan Zhang, Yiming Dong et al.

Chronos-2 is a pretrained universal forecasting model employing group attention, supporting zero-shot predictions for univariate, multivariate, and covariate-informed tasks, achieving SOTA results.

Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken et al.

DLER employs batch normalization and dynamic sampling to optimize RL, reducing output length by over 70% while surpassing baseline accuracy.

Shih-Yang Liu, Xin Dong, Ximing Lu et al.

SPA framework uses self-play finetuning to internalize world models, boosting RL performance in OOD environments by 2x on Sokoban.

Shiqi Chen, Tongyao Zhu, Zian Wang et al.

AEPO introduces dynamic entropy balancing and gradient regulation, enhancing stability and exploration in multi-turn web agent RL with only 1K samples.

Guanting Dong, Licheng Bao, Zhongyuan Wang et al.

SHaRe-SSM excels in ultra-long sequences with 52.1x energy efficiency improvement.

Kartikay Agrawal, Abhijeet Vikram, Vedant Sharma et al.