conftrace_

alignment tax

4 papers

Explore in graph

Co-occurring keywords · papers mentioning both

large language model (2) reinforcement learning with human feedback (1) controllable preference optimization (1) 3h desideratum (1) weak-to-strong decoding (1) transformer layer (1) preference score (1) multi-objective alignment (1) human preference (1) feature diversity (1)

Papers

Well Begun is Half Done: Low-resource Preference Alignment by Weak-to-Strong Decoding ACL 2025 Controllable Preference Optimization: Toward Controllable Multi-Objective Alignment EMNLP 2024 Disperse-Then-Merge: Pushing the Limits of Instruction Tuning via Alignment Tax Reduction ACL 2024 Mitigating the Alignment Tax of RLHF EMNLP 2024