README.md

🧑‍🎓 Weekly Paper Reading Group

This repository contains the summaries of papers related to the alignment problem in Natural Language Processing (NLP), and discussions with Kyoungwhan Mheen.
All the summaries and discussions are either in Korean (한국어) or English.

Description

Objective

This covers several papers related to the alignment problem and the methods such as instruction tuning and Reinforcement Learning from Human Feedback (RLHF) that attempt to solve it.

Papers

You can click the document emoji (📄) to read the summary if available.

Title	Presented	Codes	Tag	Presenter
FLAN: Finetuned Language Models Are Zero-Shot Learners	✅ 📄	google-research/FLAN	Instruction Tuning	Yumin Seol
T0: Multitask Prompted Training Enables Zero-Shot Task Generalization		bigscience-workshop/t-zero
InstructGPT: Training language models to follow instructions with human feedback
Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback		anthropics/hh-rlhf
Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks		allenai/natural-instructions
Guess the Instruction! Flipped Learning Makes Language Models Stronger Zero-Shot Learners		seonghyeonye/Flipped-Learning
Scaling Instruction-Finetuned Language Models
Exploring the Benefits of Training Expert Language Models over Instruction Tuning		joeljang/elm