Establishes global linear convergence of entropy-regularized softmax policy gradient in continuous-space MDPs via non-uniform Polyak-Łojasiewicz inequality, extending beyond tabular cases.

Ziyue Chen, David Šiška, Lukasz Szpruch

HumanEgo uses entity-level hand-object interaction representation and flow matching to enable zero-shot robot learning from minutes of human egocentric videos, achieving 92.5% success.

Zhi Wang, Botao He, Kelin Yu et al.

Introduces an entropy-regularized discrete Bellman policy for risk-sensitive market making, with proven convergence rate O(h + λ(1+|log λ|)).

Tenghan Zhong

Proposed CFCP method improves local coverage in multi-class classification using cluster frequency.

Tomer Lavi, Bracha Shapira, Nadav Rappoport

PANDO improves multimodal AI agent efficiency via online skill distillation, achieving 58.3% success rate.

Yubo Li, Yidi Miao, Yuntian Shen et al.

Proposes a template-controlled DiD protocol to accurately measure alignment-induced activation shifts, removing formatting confounds.

Yuki Nakamura

FinCAD reduces look-ahead bias in financial backtesting via context-aware decoding.

Weixian Waylon Li, Mengyu Wang, Tiejun Ma

ECHO integrates environment observation prediction with policy optimization, doubling task success rate from 2.70% to 5.17% on Qwen3-8B in TerminalBench-2.0.

Vaishnavi Shrivastava, Piero Kauffmann, Ahmed Awadallah et al.

SAM introduces an external, state-adaptive memory system that significantly improves long-horizon reasoning, outperforming baselines with 8-10% gains on benchmarks.

Yuyang Hu, Hongjin Qian, Shuting Wang et al.

MineC2FNet enhances fine-grained segmentation of mining footprints in multispectral imagery using attentive distillation, significantly improving accuracy.

Alif Tri Handoyo, Vincent C. S. Lee, Rizka Widyarini Purwanto et al.

OpenSkillEval introduces a dynamic evaluation framework using over 600 generated tasks and 30 open-source skills, revealing that skill availability does not guarantee effective utilization or performance gains.

Jiahao Ying, Boxian Ai, Wei Tang et al.

Proposes Recursive Block-Diagonal Coupling (RBDC) for training wide vision models efficiently, reducing FLOPs by 30% while maintaining accuracy.

Maxim Henry, Adrien Deliège, Sébastien Piérard et al.

DiLaDiff combines latent continuous diffusion with discrete masking, achieving high-quality, fast language generation.

Jean-Marie Lemercier, Tomas Geffner, Karsten Kreis et al.

Proposes a sparse activation-based neural scaling law model, revealing double descent peaks and asymmetry due to sparsity effects.

John Sous, Michael Winer

Proposes latent policy gradients to predict out-of-distribution RL behavior with 89.73% accuracy.

Jason Ross Brown, Edward James Young

AssetGen employs a VecSet diffusion framework, producing high-quality textured meshes in 30s suitable for real-time rendering.

Dilin Wang, Xiaoyu Xiang, Kihyuk Sohn et al.

RankElastor combines parameterized full mixing and GLU P-FFNs, improving AUC by over 0.001 on Criteo and Avazu versus the strongest baseline.

Guoming Li, Shangyu Zhang, Junwei Pan et al.

UfM* employs Gaussian mixture models for single-inference multiview disagreement-based depth uncertainty estimation, achieving high accuracy with minimal energy.

Soumya Sudhakar, Sertac Karaman, Vivienne Sze

GEMQ uses global linear programming and efficient router fine-tuning for mixed-precision quantization of MoE LLMs, reducing memory usage.

Jianing Deng, Song Wang, Dongwei Wang et al.

CoMoGen employs MaskAdapter and LoRA to enable controllable video synthesis from binary mask sequences, enhancing motion fidelity.

Adil Meric, Lin Geng Foo, Mert Kiray et al.