<Pipeline, Rationale> PINTO: Faithful Language Reasoning Using Prompt-Generated Rationales (2023.04)

<Pipeline, Rationale> PINTO: Faithful Language Reasoning Using Prompt-Generated Rationales (2023.04)

관심있는 NLP 논문을 읽어보고 간단히 정리했습니다. 혹시 부족하거나 잘못된 내용이 있다면 댓글 부탁드립니다 ️ usechatgpt init success [University of Southern California] - PINTO: prompt-based learning을 통해 rationalize하고, counterfactual regularization을 통해 faithfully reason하는 LM pipleine - 중간 사이즈의 LM을 frozen하고 rationale을 생성할 수 있도록 prompting → 생성된 rationale을 컨텍스트로 이해하고 fine-tuning 1. Introduction 굉장히 재밌는 표현인 것 같은데, reasoning tasks에 대해서 LM의 파라..

원문링크 : <Pipeline, Rationale> PINTO: Faithful Language Reasoning Using Prompt-Generated Rationales (2023.04)

 등록된 다른 글

Introduction to Programming Frameworks

Introduction to Programming Frameworks

Analysis of Algorithms(3) : Mathematical Models

Analysis of Algorithms(3) : Mathematical Models

<LK Lab, Evaluation> [FLASK] Fine-Grained Language Model Evaluation Based on Alignment Skill Sets (2023.10)

<LK Lab, Evaluation> [FLASK] Fine-Grained Language Model Evaluation Based on Alignment Skill Sets (2023.10)

[BOJ] 11404 : 플로이드 [그래프이론](Python)

[BOJ] 11404 : 플로이드 [그래프이론](Python)

[BOJ] 5525 : IOIOI [문자열](Python)

[BOJ] 5525 : IOIOI [문자열](Python)

Jacobians - vectors of derivatives

Jacobians - vectors of derivatives

[BOJ] 11053 : 가장 긴 증가하는 부분 수열 [DP](Python)

[BOJ] 11053 : 가장 긴 증가하는 부분 수열 [DP](Python)

<Multi-modal> BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models

<Multi-modal> BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models