paper-with-me

홈 › Papers

Continual Learning for Seq2Seq Generations with Transformer Calibration

2022-01-16 · ACL ARR January 2022 1 · Anonymous

Conventional NLP generation models are trained offline with a given dataset for a particular task, which is referred to as isolated learning. Research on sequence-to-sequence language generation aims to study continual learning model to constantly learning from sequentially encountered tasks. However, continual learning studies often suffer from catastrophic forgetting, a persistent challenge for lifelong learning. In this paper, we present a novel NLP transformer model which attempts to mitigate catastrophic forgetting in online continual learning from a new perspective, i.e., attention calibration. We model the attention in the transformer as a calibrated unit in a general formulation, where the attention calibration could give benefits to balance the stability and plasticity of continual learning algorithms through influencing both their forward inference path and backward optimization path. Our experiments, paraphrase generation, show that this work outperforms SOTA models by a considerable margin and remedy the forgetting greatly.

📄 PDF Abstract BibTeX

Code (0)

등록된 구현이 없습니다.

Tasks

Continual LearningLifelong learningParaphrase GenerationText Generation

Similar Papers 제목 키워드 기반

Calibration, Entropy Rates, and Memory in Language Models

2019-06-11 · ICML 2020 1 · Mark Braverman, Xinyi Chen, Sham M. Kakade, Karthik Narasimhan 외

Building accurate language models that capture meaningful long-term dependencies is a core challenge in natural language processing. Towards this end, we present a calibration-based approach to measure long-term discrepa…

Atomic Calibration of LLMs in Long-Form Generations

2024-10-17 · Caiqi Zhang, Ruihan Yang, Zhisong Zhang, Xinting Huang 외

Large language models (LLMs) often suffer from hallucinations, posing significant challenges for real-world applications. Confidence calibration, which estimates the underlying uncertainty of model predictions, is essent…

Form

Linguistic Calibration of Long-Form Generations

2024-03-30 · Neil Band, Xuechen Li, Tengyu Ma, Tatsunori Hashimoto

Language models (LMs) may lead their users to make suboptimal downstream decisions when they confidently hallucinate. This issue can be mitigated by having the LM verbally convey the probability that its claims are corre…

Decision MakingFormQuestion Answering

Mitigating Forgetting in Online Continual Learning with Neuron Calibration

2021-12-01 · NeurIPS 2021 12 · Haiyan Yin, Peng Yang, Ping Li

Inspired by human intelligence, the research on online continual learning aims to push the limits of the machine learning models to constantly learn from sequentially encountered tasks, with the data from each task being…

Continual Learning

Mitigating Forgetting in Online Continual Learning with Neuron Calibration

2021-05-21 · NeurIPS 2021 12 · Haiyan Yin, Peng Yang, Ping Li

Inspired by human intelligence, the research on online continual learning aims to push the limits of the machine learning models to constantly learn from sequentially encountered tasks, with the data from each task being…

Continual Learning