Papers with test-time weak-to-strong alignment
Reward-Shifted Speculative Sampling Is An Efficient Test-Time Weak-to-Strong Aligner (2025.emnlp-main)
Copied to clipboard
| Challenge: | Recent research has focused on test-time alignment, where additional compute is allocated during inference to enhance LLM safety and reasoning capabilities. |
| Approach: | They propose a reward-shifted speculative sampling algorithm that aligns a draft model with human preferences while the target model remains unchanged. |
| Outcome: | The proposed algorithm achieves superior gold reward scores at a significantly reduced inference cost in test-time weak-to-strong alignment experiments. |