…Daniel Peñaherrera , , Abstract Scalable inference-time annealing method uses flow-based models with energy-based surrogates to efficiently sample Boltzmann distributions without costly divergence calculations. Generated by Qwen/Qwen2.5-Coder-32B…
…The central idea of dMoE is to aggregate token-level expert distributions within each block into a unified block-level expert distribution , which is then used to guide expert routing in a…
…While WMs can model distributions over futures, policy evaluation and improvement typically rely on nominal imaginations, which can miss high-impact outcomes of robot actions unless prohibitively many samples are drawn. To…
…effectiveness depending on teacher choice, loss formulation, and instance-specific privileged information availability, with identified failure mechanisms including distribution mismatch, optimization instability, and PI-free policy learning. AI-generated summary On-policy…
…AI-generated summary While self-supervised pretraining has reduced vision systems' reliance on synthetic data, simulation remains an indispensable tool for closed-loop optimization and rigorous out-of-distribution (OOD) evaluation. However…
…Depth provides geometric context that is invariant to lighting variations, while luminance and semantic features offer explicit guidance on brightness distribution and scene understanding. Modalities are extracted at multiple scales and fused…
…In non-stationary environments, the diverse reasoning distributions of source models often evolve unpredictably, transmitting systematic biases and drift to the target model. To address this, we formulate multi-source reasoning alignment…
…Learning Accurate Hierarchical Low-Rank Representations for LLM Fine-Tuning Published on May 8 Submitted by Ionut Modoranu on May 11 IST Austria Distributed Algorithms and Systems Lab Authors: , , Abstract MatryoshkaLoRA introduces…
…Crucially, beyond its conventional role of manifold lifting (i.e., moving data off low-dimensional manifolds), injecting Gaussian noise facilitates domain harmonization by implicitly aligning feature distributions across domains, a property particularly…
…However, a recent study observes that the noise domain constructed from simple distributions (*e.g.*, Gaussian distributions) can serve as a surrogate source domain in the semi-supervised setting, where only a…