arXiv:2610. 41588v1 Announce Type: cross Abstract: Direct Preference Optimization (DPO) treats every reported comparison as correct, yet preference data contain reversed, tied or otherwise noisy comparisons.
arXiv:2607. 01715v1 Announce Type: new Abstract: Existing robust preference optimization for language-model alignment mainly studies pairwise supervision and places robustness at the dataset, prompt, or preference-pair level.
arXiv:2510. 05342v2 Announce Type: replace-cross Abstract: Direct Preference Optimization (DPO) has emerged as a simple and effective method for aligning large language models.
arXiv:2607. 16240v1 Announce Type: cross Abstract: Direct Alignment Algorithms (DAAs) such as DPO have become a common way to post-train and align LLMs with human preferences.
arXiv:2611.07375v1 Announce Type: new Abstract: Direct Preference Optimization (DPO) is increasingly used in settings where pairwise preferences coexist with pointwise criteria such as response lengt...
The paper introduces GAW-PO, a gradient‑aligned token reweighting technique for Direct Preference Optimization (DPO). It assigns weaker penalties to rejected tokens whose gradients align with preferred behavior, while maintaining stronger penalties for conflicting tokens. Across 11 benchmarks in mathematics, reasoning, coding, and question answering, GAW‑PO outperforms standard DPO and other baselines, and remains robust when the DPO regularization parameter is reduced.
arXiv:2609.38860v1 Announce Type: cross Abstract: Learning from human preferences is central to large language model (LLM) alignment, but human preference annotation is costly. Active preference lear...
The paper investigates whether accuracy and order sensitivity in large language model (LLM) multiple‑choice question (MCQ) evaluation move together. Using proprietary and open‑weight models on MMLU and ARC‑Challenge, the authors measure order sensitivity with flip rate and recall standard deviation (RStd) and find that these metrics often diverge from accuracy and from each other. They conclude that accuracy alone is insufficient and recommend reporting both flip rate and RStd alongside accuracy.
arXiv:2607. 25136v1 Announce Type: new Abstract: Research on preference optimization often varies the training objective while holding the data fixed.
arXiv:2604. 27733v2 Announce Type: replace Abstract: Aligning Large Language Models (LLMs) with human intent, whether through explicit reward modeling or direct methods such as DPO, fundamentally relies on minimizing a surrogate loss as a proxy for the true pairwise ranking objective.
The paper demonstrates that a preference‑optimization objective can learn to distinguish reliable from unreliable sources by installing a prior‑dependent reliability switch. By training on data where a source’s stated reliability is paired with its answer, the model learns to flip its response only when the stated reliability exceeds a threshold that grows with the model’s prior. Experiments on Qwen2.5‑7B‑Instruct and Llama‑3.1‑8B show that this switch generalizes to unseen reliability values and follows stated reliability over role prestige, whereas supervised imitation fails to learn it.
The paper introduces conservative contextual dueling bandits, which enforce that each displayed item’s cumulative score remains above a specified fraction of a trusted baseline’s score. It presents algorithms—MaxInP-BC for linear utilities and BTIGW-BC for general utilities—that guarantee safety with high probability while achieving sublinear regret. Experiments show these methods satisfy the safety constraint more reliably than unconstrained counterparts, with only a moderate increase in regret.
The paper introduces First-Order In-Context Shapley, an inference‑only method that evaluates data utility for large language model alignment by computing the first‑order term of an in‑context cooperative game over semantic k‑NN neighborhoods in O(k) forward passes. It projects log‑likelihood shifts onto a directed influence graph, yielding three signals—Exemplifying Advantage, Exemplified Advantage, and One‑of‑a‑Kind Score—to identify high‑utility demonstrations and rare prompts. The approach reduces compute by 19× compared to Monte‑Carlo Shapley, filters 1–5% of records, and improves error detection and reward accuracy in DPO training and benchmark auditing. whyItMatters":"By enabling scalable, inference‑only data auditing, the method addresses label noise and safety inversions that hinder post‑training alignment of large language models, leading to measurable improvements in model reliability and reward accuracy."