OmniVerifier-M1 employs symbolic bounding boxes and decoupled reinforcement learning to enhance visual verification accuracy, achieving 0.68 on ViVerBench.

Xinchen Zhang, Bowei Liu, Jiale Liu et al.

Proposes CAPO, a cross-annotator preference optimization method, enabling LLMs to learn and reproduce stable individual explanation behaviors, outperforming prompting and SFT.

Beiduo Chen, Pingjun Hong, Ziyun Zhang et al.

LearnWeak framework uses a stronger reference agent to identify model weaknesses, synthesizes targeted tasks, and improves small CUAs by 11.6% on average across 8 domains.

Suji Kim, Kangsan Kim, Sung Ju Hwang

FluxMem models memory as a dynamically evolving heterogeneous graph with three stages, achieving state-of-the-art results in complex reasoning and web navigation tasks.

Jizhan Fang, Buqiang Xu, Zhixian Wang et al.

BIRDNet encodes mined Boolean implication graphs into sparse, interpretable deep neural networks, achieving near state-of-the-art AUROC with 96x fewer active parameters on six biomedical datasets.

Tirtharaj Dash

This study reveals architecture-specific failure signatures in VLA models via black-box action monitoring, emphasizing the importance of architecture-matched monitors.

Krishnam Gupta

Using pymetrics data, reveals algorithmic monoculture causes racial disparities and systemic rejection in hiring.

Rishi Bommasani, Sarah H. Bana, Kathleen A. Creel et al.

MUSE-Autoskill improves task success to 68.4% via unified skill lifecycle management and cross-agent skill transfer.

Huawei Lin, Peng Li, Jie Song et al.

LocateAnything uses Parallel Box Decoding and 138M samples to significantly boost visual grounding speed and accuracy.

Shihao Wang, Shilong Liu, Yuanguo Kuang et al.

BRANE uses LLM-extracted query features to dynamically optimize retrieval configurations, achieving up to 89% cost savings on MuSiQue and others.

Melissa Z. Pan, Negar Arabzadeh, Mathew Jacob et al.

SAERL leverages Sparse Autoencoder activations to model diversity, difficulty, and quality for LLM post-training data engineering, boosting Qwen2.5-Math-1.5B accuracy by 3%.

Yi Jing, Zao Dai, Jinwu Hu et al.

Proposed GADD algorithm achieves O(polylog(ε⁻¹)) sampling complexity for uniform-rate discrete diffusion models, significantly accelerating sampling.

Yuchen Liang, Ness Shroff, Yingbin Liang

Representation-conditioned diffusion models using DINO embeddings achieve high-quality, controllable image generation on LSUN and CelebA datasets.

Nithesh Chandher Karthikeyan, Jonas Unger, Gabriel Eilertsen

FinHarness reduces ASR to 15% on FinVault with 4.7× fewer advanced judge calls via inline lifecycle safety harness

Haoxuan Jia, Yang Liu, Bin Chong et al.

Proposed Normal Guidance regularization improves attention-based MIL slice-level localization on 4M+ CT slices, outperforming baselines.

Ethan Harvey, Dennis Johan Loevlie, Michael C. Hughes

LLM-guided evolutionary co-evolution of constitutions achieves ~0.78 stable equilibrium in Public Goods Game arms race.

Ujwal Kumar, Arth Singh, Hershraj Niranjani et al.

Unified Neural Scaling Law (UNSL) models multi-dimensional scaling of deep networks, improving performance extrapolation accuracy by over 10%.

Ethan Caballero, Priyank Jaini, David Krueger et al.

Proposed DeltaDirect with MoDirect dataset boosts motion direction accuracy from 25.9% to 85.4% on synthetic domain.

Jongseo Lee, Hyuntak Lee, Sunghun Kim et al.

ConvexTok uses convex relaxations to optimize tokenisation, achieving near-optimal compression within 1% at 128k vocabulary, improving BpB significantly.

Jan Tempus, Philip Whittington, Craig W. Schmidt et al.

MotiMotion integrates VLM-based reasoning and confidence-aware control for motion-controlled video generation, outperforming baselines on MotiBench.

Lee Hsin-Ying, Hanwen Jiang, Yiqun Mei et al.