Zhuohao Yu, Chang Gao, Wenjin Yao, Yidong Wang, Wei Ye, Jindong Wang, Xing Xie, Yue Zhang, Shikun Zhang. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Instruction tuning large language models (LLMs) remains a challenging task, owing to the complexity of hyperparameter selection and the difficulty involved in evaluating the tuned models. To determine the optimal hyperparameters, an automatic, robust, and reliable evaluation benchmark is essential. However, establishing such a benchmark is not a trivial task due to the challenges associated with evaluation accuracy and privacy protection. In response to these challenges, we introduce a judge large language model, named PandaLM, which is trained to distinguish the superior model given several LLMs. PandaLM's focus extends beyond just the objective correctness of responses, which is the main focus of traditional evaluation datasets. It addresses vital subjective factors such as relative conciseness, clarity, adherence to instructions, comprehensiveness, and formality. To ensure the reliability of PandaLM, we collect a diverse human-annotated test dataset, where all contexts are generated by humans and labels are aligned with human preferences. Our findings reveal that PandaLM-7B offers a performance comparable to both GPT-3.5 and GPT-4. Impressively, PandaLM-70B surpasses their performance. PandaLM enables the evaluation of LLM to be fairer but with less cost, evidenced by significant improvements achieved by models tuned through PandaLM compared to their counterparts trained with default Alpaca's hyperparameters. In addition, PandaLM does not depend on API-based evaluations, thus avoiding potential data leakage.

This paper introduces AutoSurvey, a speedy and well-organized methodology for automating the creation of comprehensive literature surveys in rapidly evolving fields like artificial intelligence. Traditional survey paper creation faces challenges due to the vast volume and complexity of information, prompting the need for efficient survey methods. While large language models (LLMs) offer promise in automating this process, challenges such as context window limitations, parametric knowledge constraints, and the lack of evaluation benchmarks remain. AutoSurvey addresses these challenges through a systematic approach that involves initial retrieval and outline generation, subsection drafting by specialized LLMs, integration and refinement, and rigorous evaluation and iteration. Our contributions include a comprehensive solution to the survey problem, a reliable evaluation method, and experimental validation demonstrating AutoSurvey's effectiveness.

End-to-end Long Short-Term Memory (LSTM) has been successfully applied to video summarization. However, the weakness of the LSTM model, poor generalization with inefficient representation learning for inputted nodes, limits its capability to efficiently carry out node classification within user-created videos. Given the power of Graph Neural Networks (GNNs) in representation learning, we adopted the Graph Information Bottle (GIB) to develop a Contextual Feature Transformation (CFT) mechanism that refines the temporal dual-feature, yielding a semantic representation with attention alignment. Furthermore, a novel Salient-Area-Size-based spatial attention model is presented to extract frame-wise visual features based on the observation that humans tend to focus on sizable and moving objects. Lastly, semantic representation is embedded within attention alignment under the end-to-end LSTM framework to differentiate indistinguishable images. Extensive experiments demonstrate that the proposed method outperforms State-Of-The-Art (SOTA) methods.

The high redundancy among keyframes is a critical issue for the existing summarizing methods in dealing with user-created videos. To address the critical issue, we present an unsupervised learning method, Spatial Attention Model guided Bi-directional Long Short-term Memory network (Bi-LSTM), on the combination of visual and semantic features. As for the visual feature, we design a Salient-Area- Size-based spatial attention model on the observation that humans tend to focus on sizable and moving objects in videos. Moreover, the Bi-LSTM network is leveraged to exploit the semantic feature. Afterward, the Soft Selected Probability generated from the spatial attention and semantic feature is fused to obtain the final probability for keyframe selection. The reinforcement learning framework, trained by the Deep Deterministic Policy Gradient algorithm, is adopted to do unsupervised training. Extensive experiments on the SumMe and TVSum datasets demonstrate that our method outperforms the state-of-the-art methods in terms of F-score.