Inspecting the Factuality of Hallucinated Entities in Abstractive Summarization
Meng Cao
Yue Dong
State-of-the-art abstractive summarization systems often generate hallucinations ; i.e., content that is not directly inferable from the sou… (voir plus)rce text. Despite being assumed incorrect, many of the hallucinated contents are consistent with world knowledge (factual hallucinations). Including these factual hallucinations into a summary can be beneficial in providing additional background information. In this work, we propose a novel detection approach that separates factual from non-factual hallucinations of entities. Our method is based on an entity’s prior and posterior probabilities according to pre-trained and finetuned masked language models, respectively. Empirical re-sults suggest that our method vastly outperforms three strong baselines in both accuracy and F1 scores and has a strong correlation with human judgements on factuality classification tasks. Furthermore, our approach can provide insight into whether a particular hallucination is caused by the summarizer’s pre-training or fine-tuning step. 1
Inspecting the Factuality of Hallucinated Entities in Abstractive Summarization
Meng Cao
Yue Dong
State-of-the-art abstractive summarization systems often generate hallucinations ; i.e., content that is not directly inferable from the sou… (voir plus)rce text. Despite being assumed incorrect, many of the hallucinated contents are consistent with world knowledge (factual hallucinations). Including these factual hallucinations into a summary can be beneficial in providing additional background information. In this work, we propose a novel detection approach that separates factual from non-factual hallucinations of entities. Our method is based on an entity’s prior and posterior probabilities according to pre-trained and finetuned masked language models, respectively. Empirical re-sults suggest that our method vastly outperforms three strong baselines in both accuracy and F1 scores and has a strong correlation with human judgements on factuality classification tasks. Furthermore, our approach can provide insight into whether a particular hallucination is caused by the summarizer’s pre-training or fine-tuning step. 1
Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization
Kartik Ahuja
Ethan Caballero
Dinghuai Zhang
Jean-Christophe Gagnon-Audet
The invariance principle from causality is at the heart of notable approaches such as invariant risk minimization (IRM) that seek to address… (voir plus) out-of-distribution (OOD) generalization failures. Despite the promising theory, invariance principle-based approaches fail in common classification tasks, where invariant (causal) features capture all the information about the label. Are these failures due to the methods failing to capture the invariance? Or is the invariance principle itself insufficient? To answer these questions, we revisit the fundamental assumptions in linear regression tasks, where invariance-based approaches were shown to provably generalize OOD. In contrast to the linear regression tasks, we show that for linear classification tasks we need much stronger restrictions on the distribution shifts, or otherwise OOD generalization is impossible. Furthermore, even with appropriate restrictions on distribution shifts in place, we show that the invariance principle alone is insufficient. We prove that a form of the information bottleneck constraint along with invariance helps address key failures when invariant features capture all the information about the label and also retains the existing success when they do not. We propose an approach that incorporates both of these principles and demonstrate its effectiveness in several experiments.
Issue Link Label Recovery and Prediction for Open Source Software
Alexander Nicholson
Guo Jin L.C.
Modern open source software development heavily relies on the issue tracking systems to manage their feature requests, bug reports, tasks, a… (voir plus)nd other similar artifacts. Together, those “issues” form a complex network with links to each other. The heterogeneous character of issues inherently results in varied link types and therefore poses a great challenge for users to create and maintain the label of the link manually. The goal of most existing automated issue link construction techniques ceases with only examining the existence of links between issues. In this work, we focus on the next important question of whether we can assess the type of issue link automatically through a data-driven method. We analyze the links between issues and their labels used the issue tracking system for 66 open source projects. Using three projects, we demonstrate promising results when using supervised machine learning classification for the task of link label recovery with careful model selection and tuning, achieving F1 scores of between 0.56-0.70 for the three studied projects. Further, the performance of our method for future link label prediction is convincing when there is sufficient historical data. Our work signifies the first step in systematically manage and maintain issue links faced in practice.
Learning Neural Generative Dynamics for Molecular Conformation Generation
Minkai Xu
Shitong Luo
Jian Peng
We study how to generate molecule conformations (i.e., 3D structures) from a molecular graph. Traditional methods, such as molecular dynamic… (voir plus)s, sample conformations via computationally expensive simulations. Recently, machine learning methods have shown great potential by training on a large collection of conformation data. Challenges arise from the limited model capacity for capturing complex distributions of conformations and the difficulty in modeling long-range dependencies between atoms. Inspired by the recent progress in deep generative models, in this paper, we propose a novel probabilistic framework to generate valid and diverse conformations given a molecular graph. We propose a method combining the advantages of both flow-based and energy-based models, enjoying: (1) a high model capacity to estimate the multimodal conformation distribution; (2) explicitly capturing the complex long-range dependencies between atoms in the observation space. Extensive experiments demonstrate the superior performance of the proposed method on several benchmarks, including conformation generation and distance modeling tasks, with a significant improvement over existing generative models for molecular conformation sampling.
Learning Robust State Abstractions for Hidden-Parameter Block MDPs
Amy Zhang
Shagun Sodhani
Learning Task Decomposition with Ordered Memory Policy Network
Yuchen Lu
Yikang Shen
Siyuan Zhou
Joshua B. Tenenbaum
Chuang Gan
Many complex real-world tasks are composed of several levels of sub-tasks. Humans leverage these hierarchical structures to accelerate the l… (voir plus)earning process and achieve better generalization. In this work, we study the inductive bias and propose Ordered Memory Policy Network (OMPN) to discover subtask hierarchy by learning from demonstration. The discovered subtask hierarchy could be used to perform task decomposition, recovering the subtask boundaries in an unstruc-tured demonstration. Experiments on Craft and Dial demonstrate that our modelcan achieve higher task decomposition performance under both unsupervised and weakly supervised settings, comparing with strong baselines. OMPN can also bedirectly applied to partially observable environments and still achieve higher task decomposition performance. Our visualization further confirms that the subtask hierarchy can emerge in our model.
Learning a Universal Template for Few-shot Dataset Generalization
Eleni Triantafillou
Richard Zemel
Vincent Dumoulin
Learning with Gradient Descent and Weakly Convex Losses
Dominic Richards
We study the learning performance of gradient descent when the empirical risk is weakly convex, namely, the smallest negative eigenvalue of … (voir plus)the empirical risk's Hessian is bounded in magnitude. By showing that this eigenvalue can control the stability of gradient descent, generalisation error bounds are proven that hold under a wider range of step sizes compared to previous work. Out of sample guarantees are then achieved by decomposing the test error into generalisation, optimisation and approximation errors, each of which can be bounded and traded off with respect to algorithmic parameters, sample size and magnitude of this eigenvalue. In the case of a two layer neural network, we demonstrate that the empirical risk can satisfy a notion of local weak convexity, specifically, the Hessian's smallest eigenvalue during training can be controlled by the normalisation of the layers, i.e., network scaling. This allows test error guarantees to then be achieved when the population risk minimiser satisfies a complexity assumption. By trading off the network complexity and scaling, insights are gained into the implicit bias of neural network scaling, which are further supported by experimental findings.
Machine Learning for Combinatorial Optimization: a Methodological Tour d'Horizon
Andrea Lodi
Antoine Prouvost
MICo: Improved representations via sampling-based state similarity for Markov decision processes
Tyler Kastner
Mark Rowland
We present a new behavioural distance over the state space of a Markov decision process, and demonstrate the use of this distance as an effe… (voir plus)ctive means of shaping the learnt representations of deep reinforcement learning agents. While existing notions of state similarity are typically difficult to learn at scale due to high computational cost and lack of sample-based algorithms, our newly-proposed distance addresses both of these issues. In addition to providing detailed theoretical analyses, we provide empirical evidence that learning this distance alongside the value function yields structured and informative representations, including strong results on the Arcade Learning Environment benchmark.
MICo: Learning improved representations via sampling-based state similarity for Markov decision processes
Tyler Kastner
Mark Rowland
We present a new behavioural distance over the state space of a Markov decision process, and demonstrate the use of this distance as an eff… (voir plus)ective means of shaping the learnt representations of deep reinforcement learning agents. While existing notions of state similarity are typically difficult to learn at scale due to high computational cost and lack of sample-based algorithms, our newly-proposed distance addresses both of these issues. In addition to providing detailed theoretical analysis