Discriminative Adversarial Search for Abstractive Summarization (by Thomas Scialom)
Speaker: Thomas Scialom:
Twitter: https://twitter.com/ThomasScialom
◾ Talk Description
In this talk, Thomas Scialom will introduce a novel adversarial approach to improve beam search for sequence generation, and apply it to Abstractive Summarization: Discriminative Adversarial Search (DAS). Inspired by Generative Adversarial Networks (GANs), wherein a discriminator is used to improve the generator, our method differs from GANs in that the generator parameters are not updated at training time and the discriminator is only used to drive sequence generation at inference time.
Note: This paper was accepted at ICML 2020.
◾ Authors: Thomas Scialom, Paul-Alexis Dray, Sylvain Lamprier, Benjamin Piwowarski, Jacopo Staiano
◾ Link: https://arxiv.org/abs/2002.10375
◾ About dair.ai
Website: https://dair.ai/
GitHub: https://github.com/dair-ai
Twitter: https://twitter.com/dair_ai
Newsletter: https://dair.ai/newsletter/
Slack: https://join.slack.com/t/dairai/shared_invite/zt-dv2dwzj7-F9HT047jIGkunNKv88lQ~g
Видео Discriminative Adversarial Search for Abstractive Summarization (by Thomas Scialom) канала Elvis Saravia
Twitter: https://twitter.com/ThomasScialom
◾ Talk Description
In this talk, Thomas Scialom will introduce a novel adversarial approach to improve beam search for sequence generation, and apply it to Abstractive Summarization: Discriminative Adversarial Search (DAS). Inspired by Generative Adversarial Networks (GANs), wherein a discriminator is used to improve the generator, our method differs from GANs in that the generator parameters are not updated at training time and the discriminator is only used to drive sequence generation at inference time.
Note: This paper was accepted at ICML 2020.
◾ Authors: Thomas Scialom, Paul-Alexis Dray, Sylvain Lamprier, Benjamin Piwowarski, Jacopo Staiano
◾ Link: https://arxiv.org/abs/2002.10375
◾ About dair.ai
Website: https://dair.ai/
GitHub: https://github.com/dair-ai
Twitter: https://twitter.com/dair_ai
Newsletter: https://dair.ai/newsletter/
Slack: https://join.slack.com/t/dairai/shared_invite/zt-dv2dwzj7-F9HT047jIGkunNKv88lQ~g
Видео Discriminative Adversarial Search for Abstractive Summarization (by Thomas Scialom) канала Elvis Saravia
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
Microsoft introduces Phi-3! #ai #llms #microsoftKeep Learning ML (Session 1) | DSV, CompLex, Modern tools for emotionsExploring Capabilities of Long-Context LLMsLlama 3 is here! | First impressions and thoughtsTraining an LLM to effectively use information retrievalQuestion Understanding: COVID-Q: 1,600+ Questions about COVID-19101 ways to solve search (by Pratik Bhavsar)Best Practices and Lessons Learned on Synthetic Data for Language Models #ai #machinelearning #genaiTLDR Generation of Scientific Documents | ML Interview #1 with Isabel CacholaDive into Deep Learning (Study Group): Deep Learning Computation with PyTorch | Session 5Understanding LLM SettingsState-of-the-art open-source LLM judges #ai #machinelearning #gpt4Keep Learning ML #3 | Contrastively Trained Structured World ModelsLearn about LLMs in this NEW course #ai #chatgpt #engineeringHow I read and annotate ML papers[LLM NEWS] AlphaFold 3, xLSTM, OpenAI's Model Spec, DeepSeek-V2, OpenDevin CodeAct 1.0RAG Faithfulness #llms #ai #gpt4Dive into Deep Learning (Study Group): Multilayer Perceptrons | Session 4Reasoning with Intermediate Revision and Search with LLMs #chatgpt #ai #llms #science #programmingDive into Deep Learning (Study Group): Modern CNNs | Session 7SWE-Agent | An LLM-based Software Engineering Agent