Fei Teng

Tianchi Cai, Xierui Song, Jiyan Jiang et al.

Aligning language models to human expectations, e.g., being helpful and harmless, has become a pressing challenge for large language models. A typical alignment procedure consists of supervised fine-tuning and preference learning. Most preference learning methods, such as RLHF and DPO, depend on pairwise preference data, which inadequately address scenarios where human feedback is point-wise, leading to potential information loss and suboptimal performance. Addressing this gap, we introduce Point-wise Direct Preference Optimization, a novel preference learning method designed to harness point-wise feedback effectively. Our work also uncovers a novel connection between supervised fine-tuning and point-wise preference learning, culminating in Unified Language Model Alignment, a single-step method that unifies the alignment with human demonstrations and point-wise preferences. Extensive experiments on point-wise preference datasets with binary or continuous labels validate the effectiveness of our methods. Our code and a new dataset with high-quality demonstration samples on harmlessness are released.

Yubo Wang, Haoyang Li, Fei Teng et al.

Text classification is a fundamental task in data mining, pivotal to various applications such as tabular understanding and recommendation. Although neural network-based models, such as CNN and BERT, have demonstrated remarkable performance in text classification, their effectiveness heavily relies on abundant labeled training data. This dependency makes these models less effective in dynamic few-shot text classification, where labeled data is scarce, and new target labels frequently appear based on application needs. Recently, large language models (LLMs) have shown promise due to their extensive pretraining and contextual understanding ability. Current approaches provide LLMs with text inputs, candidate labels, and additional side information (e.g., descriptions) to classify texts. However, their effectiveness is hindered by the increased input size and the noise introduced through side information processing. To address these limitations, we propose a graph-based online retrieval-augmented generation framework, namely GORAG, for dynamic few-shot text classification. Rather than treating each input independently, GORAG constructs and maintains a weighted graph by extracting side information across all target texts. In this graph, text keywords and labels are represented as nodes, with edges indicating the correlations between them. To model these correlations, GORAG employs an edge weighting mechanism to prioritize the importance and reliability of extracted information and dynamically retrieves relevant context using a minimum-cost spanning tree tailored for each text input. Empirical evaluations demonstrate that GORAG outperforms existing approaches by providing more comprehensive and precise contextual information.