Reward fine-tuning aims to update a pre-trained flow-based generative model to improve the downstream reward of its generated samples. Existing methods typically formulate this problem as sampling from a reward-tilted distribution, the solution to a KL-regularized reward-maximization problem. Here, we introduce an opti...
Abbas Mammadov, Jerry Huang, Justin Lin et al.· arXiv.org· 0 citations
SafeMoE, a Mixture-of-Experts (MoE) framework that isolates unsafe knowledge into domain-specific Low-Rank Adapters (LoRA experts) trained exclusively on harmful corpora, is introduced, proposing a dialectical approach to alignment, positing that unsafe data encodes rich, domain specific knowledge critical for nuanced,...
M. Hashemzadeh, Jerry Huang, Minseon Kim et al.· arXiv.org· 0 citations
FinancialAuditBench is introduced, a benchmark for evaluating agents on financial statement audit tasks, along with a framework for systematically generating synthetic engagements for model evaluation and training in privacy-sensitive domains.
Jerry Huang, Sarvesh Babu, Matt Van Buren et al.· arXiv.org· 0 citations
It is argued that accelerated samplers such as flow maps are essential infrastructure for efficient post-training, and that the dominant KL-regularized formulation is only one of many choices worth revisiting.
Abbas Mammadov, Jerry Huang, Justin Lin et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.