Picture for Jy-yong Sohn

Jy-yong Sohn

Buffer-based Gradient Projection for Continual Federated Learning

Add code
Sep 03, 2024
Viaarxiv icon

Memorization Capacity for Additive Fine-Tuning with Small ReLU Networks

Add code
Aug 01, 2024
Viaarxiv icon

Improving Multi-lingual Alignment Through Soft Contrastive Learning

Add code
May 28, 2024
Viaarxiv icon

Aligning Large Language Models for Enhancing Psychiatric Interviews through Symptom Delineation and Summarization

Add code
Mar 26, 2024
Viaarxiv icon

ERD: A Framework for Improving LLM Reasoning for Cognitive Distortion Classification

Add code
Mar 21, 2024
Viaarxiv icon

Re-Ex: Revising after Explanation Reduces the Factual Errors in LLM Responses

Add code
Feb 27, 2024
Figure 1 for Re-Ex: Revising after Explanation Reduces the Factual Errors in LLM Responses
Figure 2 for Re-Ex: Revising after Explanation Reduces the Factual Errors in LLM Responses
Figure 3 for Re-Ex: Revising after Explanation Reduces the Factual Errors in LLM Responses
Figure 4 for Re-Ex: Revising after Explanation Reduces the Factual Errors in LLM Responses
Viaarxiv icon

Analysis of Using Sigmoid Loss for Contrastive Learning

Add code
Feb 20, 2024
Viaarxiv icon

Can Separators Improve Chain-of-Thought Prompting?

Add code
Feb 16, 2024
Figure 1 for Can Separators Improve Chain-of-Thought Prompting?
Figure 2 for Can Separators Improve Chain-of-Thought Prompting?
Figure 3 for Can Separators Improve Chain-of-Thought Prompting?
Figure 4 for Can Separators Improve Chain-of-Thought Prompting?
Viaarxiv icon

Mini-Batch Optimization of Contrastive Loss

Add code
Jul 12, 2023
Viaarxiv icon

Looped Transformers as Programmable Computers

Add code
Jan 30, 2023
Viaarxiv icon