conftrace_

direct preference optimization (dpo)

2 papers

Explore in graph

Co-occurring keywords · papers mentioning both

gsm8k (1) llm reasoning capability (1) rollout sampling (1) token-level contrastive estimation (1) jailbreak (1) toxicity (1) critical token (1) alignment algorithm (1) pre-trained language model (1) interpretability (1)

Papers

Critical Tokens Matter: Token-Level Contrastive Estimation Enhances LLM’s Reasoning Capability ICML 2025 A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicity ICML 2024