Kai Ye

Kai Ye, Hongyi Zhou, Jin Zhu et al.

Reinforcement learning from human feedback (RLHF) has emerged as a key technique for aligning the output of large language models (LLMs) with human preferences. To learn the reward function, most existing RLHF algorithms use the Bradley-Terry model, which relies on assumptions about human preferences that may not reflect the complexity and variability of real-world judgments. In this paper, we propose a robust algorithm to enhance the performance of existing approaches under such reward model misspecifications. Theoretically, our algorithm reduces the variance of reward and policy estimators, leading to improved regret bounds. Empirical evaluations on LLM benchmark datasets demonstrate that the proposed algorithm consistently outperforms existing methods, with 77-81% of responses being favored over baselines on the Anthropic Helpful and Harmless dataset. The code is available at https:// github.com/ VRPO/ VRPO.

Erhan Xu, Kai Ye, Hongyi Zhou et al.

This paper studies reinforcement learning from human feedback (RLHF) for aligning large language models with human preferences. While RLHF has demonstrated promising results, many algorithms are highly sensitive to misspecifications in the underlying preference model (e.g., the Bradley-Terry model), the reference policy, or the reward function, resulting in undesirable fine-tuning. To address model misspecification, we propose a doubly robust preference optimization algorithm that remains consistent when either the preference model or the reference policy is correctly specified (without requiring both). Our proposal demonstrates superior and more robust performance than state-of-the-art algorithms, both in theory and in practice. The code is available at https://github.com/DRPO4LLM/DRPO4LLM

Ang Jia, Ming Fan, Xi Xu et al.

The great influence of Bitcoin has promoted the rapid development of blockchain-based digital currencies, especially the altcoins, since 2013. However, most altcoins share similar source codes, resulting in concerns about code innovations. In this paper, an empirical study on existing altcoins is carried out to offer a thorough understanding of various aspects associated with altcoin innovations. Firstly, we construct the dataset of altcoins, including source code repositories, GitHub fork relations, and market capitalizations (cap). Then, we analyze the altcoin innovations from the perspective of source code similarities. The results demonstrate that more than 85% of altcoin repositories present high code similarities. Next, a temporal clustering algorithm is proposed to mine the inheritance relationship among various altcoins. The family pedigrees of altcoin are constructed, in which the altcoin presents similar evolution features as biology, such as power-law in family size, variety in family evolution, etc. Finally, we investigate the correlation between code innovations and market capitalization. Although we fail to predict the price of altcoins based on their code similarities, the results show that altcoins with higher innovations reflect better market prospects.