Harnessing Black-Box Control to Boost Commonsense in LM’s Generation
Date Presented: 2/1/2024
Speaker: Yufei Tian, UCLA
Abstract:
Large language models like Alpaca and GPT-3 generate coherent texts but sometimes lack commonsense, yet improving their commonsense via fine-tuning is resource expensive in terms of both data and computation. In this talk, I’ll present BOOST, a resource-efficient framework that steers a frozen Pre-Trained Language Model (PTLM) towards more reasonable outputs. This involves creating an interpretable and reference-free evaluator that assigns a sentence with a commonsensical score which grounds the sentence to a dynamic commonsense knowledge base. Using this evaluator as a guide, we extend the NADO controllable generation method to train an auxiliary head that improves the PTLM’s output. Our framework was tested on various language models, including GPT-2, Flan-T5, and Alpaca-based models. On two constrained concept-to-sentence benchmarks, human evaluation results show that BOOST consistently generates the most commonsensical content. Finally, I will demonstrate how ChatGPT outputs are different from and sometimes less favored than our outputs.
Speaker's bio:
Yufei Tian is a CS PhD student at UCLA advised by Prof. Nanyun (Violet) Peng. Her research is centered around creative and controllable text generation, machine reasoning and its interaction with cognitive science, as well as designing evaluation metrics for open-ended NLG tasks. She is supported by the UCLA-Amazon fellowship program.
Видео Harnessing Black-Box Control to Boost Commonsense in LM’s Generation канала USC Information Sciences Institute
Speaker: Yufei Tian, UCLA
Abstract:
Large language models like Alpaca and GPT-3 generate coherent texts but sometimes lack commonsense, yet improving their commonsense via fine-tuning is resource expensive in terms of both data and computation. In this talk, I’ll present BOOST, a resource-efficient framework that steers a frozen Pre-Trained Language Model (PTLM) towards more reasonable outputs. This involves creating an interpretable and reference-free evaluator that assigns a sentence with a commonsensical score which grounds the sentence to a dynamic commonsense knowledge base. Using this evaluator as a guide, we extend the NADO controllable generation method to train an auxiliary head that improves the PTLM’s output. Our framework was tested on various language models, including GPT-2, Flan-T5, and Alpaca-based models. On two constrained concept-to-sentence benchmarks, human evaluation results show that BOOST consistently generates the most commonsensical content. Finally, I will demonstrate how ChatGPT outputs are different from and sometimes less favored than our outputs.
Speaker's bio:
Yufei Tian is a CS PhD student at UCLA advised by Prof. Nanyun (Violet) Peng. Her research is centered around creative and controllable text generation, machine reasoning and its interaction with cognitive science, as well as designing evaluation metrics for open-ended NLG tasks. She is supported by the UCLA-Amazon fellowship program.
Видео Harnessing Black-Box Control to Boost Commonsense in LM’s Generation канала USC Information Sciences Institute
Комментарии отсутствуют
Информация о видео
2 февраля 2024 г. 21:00:22
00:54:34
Другие видео канала
