Cross-Lingual Alignment for Decoder-Only Models using MoE Routers

Cross-lingual MoE router alignment improves multilingual LLM performance by aligning router outputs across languages instead of hidden states.

Lucas Bandarkar, Clark Peng, Ahmed Haj Ahmed, Aditi Khandelwal and 1 more

Published Oct 1, 2026 · 0 citations · ▲ 1 on Hugging Face · Code

OPUS: Towards Efficient and Principled Data Selection in Large Language Model Pre-training in Every Iteration

OPUS defines optimizer-induced update-space data utility for dynamic LLM pre-training selection, outperforming full-scale baselines with minimal overhead.

Shaobo Wang, Xuan Ouyang, Tianyi Xu, Yuzheng Hu and 8 more

Published Feb 5, 2026 · 0 citations · ▲ 354 on Hugging Face

Aligning Few-Step Generative Model via Amortizing Sample-Based Variational Inference

Jaewoo Lee, Hyeongyu Kang, Dohyun Kim, Kyuil Sim and 8 more

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

A Surrogate Perspective on Convergence of Fixed-Target DQN

Zichu Liu, Nneka M Okolo, Ryan D'Orazio, Danilo Vucetic and 2 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Reconstruction or Semantics? What Makes a Latent Space Useful for Robotic World Models

Nilaksh, Saurav Jha, Artem Zholus, Sarath Chandar

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Probabilistic Recursive Reasoning

Junyeob Baek, Mingyu Jo, Minsu Kim, Mengye Ren and 2 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Representation Learning Enables Scalable Multitask Deep Reinforcement Learning

Predictive representation learning combined with high-capacity value approximation drives scalable multitask RL, with the simple model-free MR.Q outperforming world-model methods across continuous control tasks.

Johan Obando Ceron, Lu Li, Scott Fujimoto, Pierre-Luc Bacon and 2 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Beyond One-Size-Fits-All: Diagnosis-Driven Online Reinforcement Learning with Offline Priors

Diagnosis-driven tension management adapts online RL to deployment-specific prior validity shifts, rejecting universal benchmarks for flexible, evidence-guided optimization.

Guozheng Ma, Lu Li, Zilin Wang, Pierre-Luc Bacon and 1 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Plausibility Is Not Prediction: Contrastive Evidence for LLM-Based Cellular Perturbation Reasoning

LLM-based cellular perturbation reasoning relies on intrinsic gene tendencies rather than true perturbation effects, and contrastive evidence organization improves prediction accuracy substantially.

XINYU YUAN, Xixian Liu, Jianan Zhao, Ya Shi Zhang and 2 more

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Sparse Koopman Autoencoders Identify Local Dynamical Regimes in Multibasin Systems

Sparse Koopman autoencoders use sparse latent supports as label-free regime indicators that identify local dynamical basins and outperform dense autoencoders in multibasin forecasting.

Aidan Li, Uday Kiran Reddy Tadipatri, Mahan Fathi, Sarath Chandar and 1 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Inverting the Bellman Equation: From $Q$-Values to World Models

Value-based agents trained on diverse reward functions implicitly encode world models, extractable via P-learning, with sufficient conditions for exact dynamics recovery and cross-goal generalization.

Alistair Letcher, Mattie Fellows, Alexander D. Goldie, Jonathan Richens and 2 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Controllable and Content Based Recommendations

CCBR builds recommendations from textual user profiles via content-derived text bottlenecks, enabling controllable multimodal steering with competitive accuracy across image, audio, and video datasets.

Firat Oncel, Jihoon Jeong, Emiliano Penaloza, Mirco Ravanelli and 2 more

Atlanta Poster Session 4, Thu, Dec 10, 4:30 PM–7:30 PM, Hall C1 · Published 2026

Phases of Muon: When Muon Eclipses SignSGD

Spectral optimizer analysis reveals three phases where Muon's SignSVD preconditions covariance differently than SignSGD.

Elliot Paquette, Noah Marshall, Lucas Benigni, Guangyuan Wang and 2 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

From Static Policies to Adaptive Priors in Offline Reinforcement Learning

Offline RL should prioritize adaptive policy priors preserving improvement capacity during online updates rather than static conservative deployment.

Tianwei Ni, Vineet Jain, Akash Karthikeyan, Pierre-Luc Bacon

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

GeneZip: Region-Aware Compression for Long Context DNA Modeling

GeneZip uses region-aware compression to achieve high base-pairs-per-token ratios, improves DNA modeling benchmarks, and enables 128K-context training on limited hardware.

Jianan Zhao, Xixian Liu, Zhihao Zhan, XINYU YUAN and 2 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026