8 accepted papers
Reinforcement finetuning (RFT) has garnered significant attention in recent years, particularly for enhancing large reasoning models such as OpenAI o1 and Deepseek R1. The appeal of RFT largely stems from its ability to refine model knowledge, better align outputs with user intent, and address chall…
In-context learning (ICL) has demonstrated significant performance improvements in transformer-based large models. This study identifies two key factors influencing ICL generalisation under complex non-i.i.d. scenario: algorithmic stability and distributional discrepancy. First, we establish a stabi…
High-quality data is essential for training large generative models, yet the vast reservoir of real data available online has become nearly depleted. Consequently, models increasingly generate their own data for further training, forming Self-consuming Training Loops (STLs). However, the empirical r…
Large generative models are increasingly trained on synthetic data from earlier generations, raising concerns about *model collapse*, a progressive performance decline consistently observed in empirical studies. However, theoretical understanding of recursive training dynamics and their failure mode…
This paper studies how to tune the stepping schedule in diffusion models, which is mostly fixed in current practice, lacking theoretical foundations and assurance of optimal performance at the chosen discretization points. In this paper, we advocate the use of adaptive time-stepping schedules and de…
This paper presents the first theoretical guarantee for Bayesian bilevel optimization (BBO) that we term for the prevalent bilevel framework combining Bayesian optimization at the outer level to tune hyperparameters, and the inner-level stochastic gradient descent (SGD) for training the model. We pr…
This paper tackles the emerging challenge of training generative models within a self-consuming loop, wherein successive generations of models are recursively trained on mixtures of real and synthetic data from previous generations. We construct a theoretical framework to rigorously evaluate how thi…
Generalization analysis of learning algorithms often builds on a critical assumption that training examples are independently and identically distributed, which is often violated in practical problems such as time series prediction. In this paper, we use algorithmic stability to study the generaliza…