Introduces TILR, a low-rank subspace method that identifies stable reasoning directions, improving model consistency by ~10% and reducing trajectory variance by 50%.

Arun Vignesh Malarkkan, Manan Roy Choudhury, Utkarsh Byahut et al.

TrafficAlign synthesizes traffic scenarios from videos, uses DSL validation, and fine-tunes LLMs for regional traffic distribution alignment, improving autonomous driving safety.

Zhi Tu, Liangkun Niu, Tianyi Zhang

Efficient spatio-temporal grounding via second-level tracking and RL verification, enhancing localization quality.

Tianshu Zhang, Yan Wang, Ji Qi et al.

EVLA fuses multimodal perception with vehicle physics via UCSE and ESRC, achieving energy-efficient driving decisions with +0.0871 score improvement.

Yuxin Liu, Zihan Chen, Haoyu Wang et al.

DLGStream introduces dual-opacity language Gaussian representation with deformation fields, achieving 43KB/frame and 60FPS in open-vocabulary free-viewpoint video streaming.

Zhihui Ke, Yuyang Liu, Xiaobo Zhou et al.

Ground4D integrates VGGT-based geometry initialization with deformable Gaussian Splatting for monocular 4D scene reconstruction and novel view synthesis, achieving state-of-the-art accuracy.

Qing Zhao, Weijian Deng, Pengxu Wei et al.

CoGS introduces a compositional Gaussian splatting framework with a six-stage optimization for monocular human-object scene reconstruction, achieving state-of-the-art fidelity.

Jerrin Bright, John Zelek

Human2Any employs object-centric interaction priors learned from human videos, enabling cross-embodiment robot manipulation without real robot demonstrations.

Shuo Cheng, Chuye Zhang, Alfred Cueva et al.

KG4VD leverages multimodal knowledge graphs and graph retrieval, achieving 44.52% accuracy on long-range document VQA benchmarks.

Yi-Cheng Wang, Chu-Song Chen

NEXT framework uses NEXT-8B to improve video recommendations, achieving +0.53% watch time and +0.51% diversity.

Yuming Liu, Hongye Yang, Harrison Zhao et al.

Proposes a self-supervised theorem discovery algorithm, autonomously generating thousands of meaningful theorems within a formal axiomatic system, enhancing proof capabilities.

Kazuki Ota, Takayuki Osa, Tatsuya Harada

COMPASS uses shared token τc to connect composition recognition and generation, trained on 389,031 images and about 1.2M VQA pairs.

Ziqi Zhou, Weize Quan, Mining Tan et al.

Obliviate effectively erases concepts in autoregressive image generation, reducing nudity detection on RAB benchmark from 91.58% to 3.15%.

Hossein Shakibania, Jonas Henry Grebe, Tobias Braun et al.

Using surrogate-gated generation and foundation-model embeddings, this study reduces evaluation calls by 90% in Bayesian materials design.

Sk Md Ahnaf Akif Alvi, Jan Janssen, Danny Perez et al.

IMCBench evaluates multimodal LLMs in image-grounded medical dialogues; Claude Opus 4.6 scores highest at 3.61.

Maria Xenochristou, Ashutosh Joshi, Korosh Vatanparvar et al.

StructSplat is a pose-free 3D Gaussian reconstruction framework achieving +5.67 dB PSNR over SOTA on DL3DV.

Jia-Chen Zhao, Beiqi Chen, Xinyang Chen et al.

PEHT integrates LoRA-enhanced Transformer with urban mobility and congestion data, achieving state-of-the-art network traffic prediction with over 90% parameter reduction.

Abdolazim Rezaei, Mehdi Sookhak, Mahboobeh Haghparast

COCOLogic-V2 evaluates visual inductive reasoning with truly hard negatives, revealing models' struggles on boundary samples.

David Steinmann, Antonia Wüst, Kristian Kersting et al.

PA-BiCoop framework improves RLBench2 tasks by 48% on average and over 50% in real-world tasks.

Bai Qicheng, Wang Ziru, Ma Teli et al.

LLawCo framework uses failure reflection to extract behavioral laws, improving multi-agent cooperation success rate by 4.5%.

Qinhong Zhou, Chuang Gan, Anoop Cherian