Ax Sungha Kim, Gawon Lee, Jusuk Lee, Jonghae Park, H. Jin Kim, Daesol Cho 6/1/2026

FLAG: Flow Policy MaxEnt-RL by Latent Augmented Guidance

arXiv paper presenting FLAG, a maximum entropy RL approach using latent augmented guidance with flow policies for high-dimensional action spaces.

Ax Yiming Ren, Yiran Xu, Zicheng Lin, Chufan Shi, Yukang Chen, Dingdong Wang, Tianhe Wu, Junjie Wang, Yujiu Yang, Yu Qiao, Ruihang Chu 6/1/2026

Smaller Models are Natural Explorers for Policy-Level Diversity in GRPO

arXiv paper showing smaller LLM models exhibit higher policy-level diversity in GRPO, improving rollout diversity without token-level noise.

Ax Zhichao Han, Mengyi Chen, Qianxiao Li 6/1/2026

Learning Permutation-invariant Macroscopic Dynamics

Learns permutation-invariant macroscopic dynamics of high-dimensional particle systems without assuming fixed ordering of microscopic degrees of freedom.

Ax Yuwei Zhang, Chengyu Dong, Shuowei Jin, Changlong Yu, Hejie Cui, Hongye Jin, Xinyang Zhang, Hamed Bonab, Colin Lockard, Jianshu Chen, Zhenyu Shi, Jingbo Shang, Xian Li, Bing Yin 6/1/2026

CoMem: Context Management with A Decoupled Long-Context Model

Introduces CoMem framework decoupling memory management from agent workflows to reduce latency overhead in long-context agentic models.

Ax Xinyang Lu, Jiabao Pan, Rachael Hwee Ling Sim, See-Kiong Ng, Anthony Kum Hoe Tung, Bryan Kian Hsiang Low 6/1/2026

De-attribute to Forget for LLM Unlearning

Proposes de-attribution method for LLM unlearning to address over-forgetting and model utility degradation when removing inappropriate training data.

Ax Jeffrey Seely, Bart{\l}omiej Cupia{\l}, Llion Jones 6/1/2026

Learning Multi-Agent Coordination via Sheaf-ADMM

Differentiable multi-agent coordination framework using sheaf-ADMM where agents solve convex subproblems and coordinate via neural encoders.

Ax Jeffrey Seely, Julian Gould 6/1/2026

Augmented Lagrangian Predictive Coding

Local learning algorithm for training deep networks via energy minimization with layer-local constraint tracking instead of backpropagation.

Ax Veronika Semmelrock, Benedetta Strizzolo, Francesco Zuccato, Gerhard Friedrich, Patrick Rodler, Konstantin Schekotihin 6/1/2026

Learning to Solve and Optimize by Evolving Code

CHECKMATE tool generates optimization algorithms via code evolution to solve combinatorial problems without expert-designed heuristics.

Ax Miltiadis Stouras, Vincent Cohen-Addad, Silvio Lattanzi, Ola Svensson 6/1/2026

Retriever Portfolios: A Principled Approach to Adaptive RAG

Method for selecting diverse retriever portfolios from large pools to adaptively handle heterogeneous RAG queries with principled query distribution coverage.

Ax Andrea Miele, Yiming Qin, Alba Carballo-Castro, Justin Deschenaux, Pascal Frossard 6/1/2026

Fixed-Point Masked Generative Modeling

Fixed-point masked generative models enable efficient parallel decoding with dynamic denoiser computation allocation per refinement step.