Papers with supervised loss
Phased Instruction Fine-Tuning for Large Language Models (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing methods to enhance pre-trained language models' ability to follow instructions are limited due to the simultaneous handling of varying instruction complexities. |
| Approach: | They propose a phased instruction fine-tuning method that posits that the transition of a pre-trained language model from simple next-word prediction to sophisticated instruction following is a gradual learning process. |
| Outcome: | The proposed method surpasses the one-off instruction fine-tuning method in win rate and validates the hypothesis of progressive alignment. |
Frictional Agent Alignment Framework: Slow Down and Don’t Break Things (2025.acl-long)
Copied to clipboard
| Challenge: | Common preference alignment methods excel in static settings, but struggle in dynamic collaborative tasks where explicit signals of interlocutor beliefs are sparse and skewed. |
| Approach: | They propose a Frictional Agent Alignment Framework to generate precise, context-aware friction that prompts deliberation and re-examination of existing evidence. |
| Outcome: | The proposed framework outperforms existing methods in producing concise, interpretable friction and in OOD generalization. |