Picture for Yijin Liu

Yijin Liu

Outdated Issue Aware Decoding for Factual Knowledge Editing

Add code
Jun 06, 2024
Viaarxiv icon

LCS: A Language Converter Strategy for Zero-Shot Neural Machine Translation

Add code
Jun 06, 2024
Viaarxiv icon

Language Generation with Strictly Proper Scoring Rules

Add code
May 29, 2024
Viaarxiv icon

Comments as Natural Logic Pivots: Improve Code Generation via Comment Perspective

Add code
Apr 11, 2024
Figure 1 for Comments as Natural Logic Pivots: Improve Code Generation via Comment Perspective
Figure 2 for Comments as Natural Logic Pivots: Improve Code Generation via Comment Perspective
Figure 3 for Comments as Natural Logic Pivots: Improve Code Generation via Comment Perspective
Figure 4 for Comments as Natural Logic Pivots: Improve Code Generation via Comment Perspective
Viaarxiv icon

Accelerating Inference in Large Language Models with a Unified Layer Skipping Strategy

Add code
Apr 10, 2024
Viaarxiv icon

Improving Translation Faithfulness of Large Language Models via Augmenting Instructions

Add code
Aug 24, 2023
Figure 1 for Improving Translation Faithfulness of Large Language Models via Augmenting Instructions
Figure 2 for Improving Translation Faithfulness of Large Language Models via Augmenting Instructions
Figure 3 for Improving Translation Faithfulness of Large Language Models via Augmenting Instructions
Figure 4 for Improving Translation Faithfulness of Large Language Models via Augmenting Instructions
Viaarxiv icon

Instruction Position Matters in Sequence Generation with Large Language Models

Add code
Aug 23, 2023
Figure 1 for Instruction Position Matters in Sequence Generation with Large Language Models
Figure 2 for Instruction Position Matters in Sequence Generation with Large Language Models
Figure 3 for Instruction Position Matters in Sequence Generation with Large Language Models
Figure 4 for Instruction Position Matters in Sequence Generation with Large Language Models
Viaarxiv icon

Towards Multiple References Era -- Addressing Data Leakage and Limited Reference Diversity in NLG Evaluation

Add code
Aug 10, 2023
Figure 1 for Towards Multiple References Era -- Addressing Data Leakage and Limited Reference Diversity in NLG Evaluation
Figure 2 for Towards Multiple References Era -- Addressing Data Leakage and Limited Reference Diversity in NLG Evaluation
Figure 3 for Towards Multiple References Era -- Addressing Data Leakage and Limited Reference Diversity in NLG Evaluation
Figure 4 for Towards Multiple References Era -- Addressing Data Leakage and Limited Reference Diversity in NLG Evaluation
Viaarxiv icon

BranchNorm: Robustly Scaling Extremely Deep Transformers

Add code
May 04, 2023
Figure 1 for BranchNorm: Robustly Scaling Extremely Deep Transformers
Figure 2 for BranchNorm: Robustly Scaling Extremely Deep Transformers
Figure 3 for BranchNorm: Robustly Scaling Extremely Deep Transformers
Figure 4 for BranchNorm: Robustly Scaling Extremely Deep Transformers
Viaarxiv icon

Towards Robust Online Dialogue Response Generation

Add code
Mar 07, 2022
Figure 1 for Towards Robust Online Dialogue Response Generation
Figure 2 for Towards Robust Online Dialogue Response Generation
Figure 3 for Towards Robust Online Dialogue Response Generation
Figure 4 for Towards Robust Online Dialogue Response Generation
Viaarxiv icon