Towards Transfer Unlearning: Empirical Evidence of Cross-Domain Bias Mitigation

Large language models (LLMs) often inherit biases from vast amounts of training corpora. Traditional debiasing methods, while effective to some extent, do not completely eliminate memorized biases and toxicity in LLMs. In this paper, we study an unlearning-based approach to debiasing in LLMs by performing gradient ascent on hate speech against minority groups, i.e., minimizing the likelihood of biased or toxic content. Specifically, we propose a mask language modeling unlearning technique, which unlearns the harmful part of the text. This method enables LLMs to selectively forget and disassociate from biased and harmful content. Experimental results demonstrate the effectiveness of our approach in diminishing bias while maintaining the language modeling abilities. Surprisingly, the results also unveil an unexpected potential for cross-domain transfer unlearning: debiasing in one bias form (e.g. gender) may contribute to mitigating others (e.g. race and religion).

Differentiable Instruction Optimization for Cross-Task Generalization

Instruction tuning has been attracting much attention to achieve generalization ability across a wide variety of tasks. Although various types of instructions have been manually created for instruction tuning, it is still unclear what kind of instruction is optimal to obtain cross-task generalization ability. This work presents instruction optimization, which optimizes training instructions with respect to generalization ability. Rather than manually tuning instructions, we introduce learnable instructions and optimize them with gradient descent by leveraging bilevel optimization. Experimental results show that the learned instruction enhances the diversity of instructions and improves the generalization ability compared to using only manually created instructions.

SciReviewGen: A Large-scale Dataset for Automatic Literature Review Generation

Automatic literature review generation is one of the most challenging tasks in natural language processing. Although large language models have tackled literature review generation, the absence of large-scale datasets has been a stumbling block to the progress. We release SciReviewGen, consisting of over 10,000 literature reviews and 690,000 papers cited in the reviews. Based on the dataset, we evaluate recent transformer-based summarization models on the literature review generation task, including Fusion-in-Decoder extended for literature review generation. Human evaluation results show that some machine-generated summaries are comparable to human-written reviews, while revealing the challenges of automatic literature review generation such as hallucinations and a lack of detailed information. Our dataset and code are available at https://github.com/tetsu9923/SciReviewGen.

Weakly Supervised Instance Segmentation using Motion Information via Optical Flow

Weakly supervised instance segmentation has gained popularity because it reduces high annotation cost of pixel-level masks required for model training. Recent approaches for weakly supervised instance segmentation detect and segment objects using appearance information obtained from a static image. However, it poses the challenge of identifying objects with a non-discriminatory appearance. In this study, we address this problem by using motion information from image sequences. We propose a two-stream encoder that leverages appearance and motion features extracted from images and optical flows. Additionally, we propose a novel pairwise loss that considers both appearance and motion information to supervise segmentation. We conducted extensive evaluations on the YouTube-VIS 2019 benchmark dataset. Our results demonstrate that the proposed method improves the Average Precision of the state-of-the-art method by 3.1.

Unsupervised Abstractive Opinion Summarization by Generating Sentences with Tree-Structured Topic Guidance

This paper presents a novel unsupervised abstractive summarization method for opinionated texts. While the basic variational autoencoder-based models assume a unimodal Gaussian prior for the latent code of sentences, we alternate it with a recursive Gaussian mixture, where each mixture component corresponds to the latent code of a topic sentence and is mixed by a tree-structured topic distribution. By decoding each Gaussian component, we generate sentences with tree-structured topic guidance, where the root sentence conveys generic content, and the leaf sentences describe specific topics. Experimental results demonstrate that the generated topic sentences are appropriate as a summary of opinionated texts, which are more informative and cover more input contents than those generated by the recent unsupervised summarization model (Bra\v{z}inskas et al., 2020). Furthermore, we demonstrate that the variance of latent Gaussians represents the granularity of sentences, analogous to Gaussian word embedding (Vilnis and McCallum, 2015).

Predictive Process Model Monitoring using Recurrent Neural Networks

The field of predictive process monitoring focuses on modelling future characteristics of running business process instances, typically by either predicting the outcome of particular objectives (e.g. completion (time), cost), or next-in-sequence prediction (e.g. what is the next activity to execute). This paper introduces Processes-As-Movies (PAM), a technique that provides a middle ground between these predictive monitoring. It does so by capturing declarative process constraints between activities in various windows of a process execution trace, which represent a declarative process model at subsequent stages of execution. This high-dimensional representation of a process model allows the application of predictive modelling on how such constraints appear and vanish throughout a process' execution. Various recurrent neural network topologies tailored to high-dimensional input are used to model the process model evolution with windows as time steps, including encoder-decoder long short-term memory networks, and convolutional long short-term memory networks. Results show that these topologies are very effective in terms of accuracy and precision to predict a process model's future state, which allows process owners to simultaneously verify what linear temporal logic rules hold in a predicted process window (objective-based), and verify what future execution traces are allowed by all the constraints together (trace-based).

Unsupervised Neural Single-Document Summarization of Reviews via Learning Latent Discourse Structure and its Ranking

This paper focuses on the end-to-end abstractive summarization of a single product review without supervision. We assume that a review can be described as a discourse tree, in which the summary is the root, and the child sentences explain their parent in detail. By recursively estimating a parent from its children, our model learns the latent discourse tree without an external parser and generates a concise summary. We also introduce an architecture that ranks the importance of each sentence on the tree to support summary generation focusing on the main review point. The experimental results demonstrate that our model is competitive with or outperforms other unsupervised approaches. In particular, for relatively long reviews, it achieves a competitive or better performance than supervised models. The induced tree shows that the child sentences provide additional information about their parent, and the generated summary abstracts the entire review.