Reducing Hallucinations in LLMs via Factuality-Aware Preference Learning
By: Sindhuja Chaduvula , Ahmed Y. Radwan , Azib Farooq and more
Potential Business Impact:
Makes AI tell the truth, not make things up.
Preference alignment methods such as RLHF and Direct Preference Optimization (DPO) improve instruction following, but they can also reinforce hallucinations when preference judgments reward fluency and confidence over factual correctness. We introduce F-DPO (Factuality-aware Direct Preference Optimization), a simple extension of DPO that uses only binary factuality labels. F-DPO (i) applies a label-flipping transformation that corrects misordered preference pairs so the chosen response is never less factual than the rejected one, and (ii) adds a factuality-aware margin that emphasizes pairs with clear correctness differences, while reducing to standard DPO when both responses share the same factuality. We construct factuality-aware preference data by augmenting DPO pairs with binary factuality indicators and synthetic hallucinated variants. Across seven open-weight LLMs (1B-14B), F-DPO consistently improves factuality and reduces hallucination rates relative to both base models and standard DPO. On Qwen3-8B, F-DPO reduces hallucination rates by five times (from 0.424 to 0.084) while improving factuality scores by 50 percent (from 5.26 to 7.90). F-DPO also generalizes to out-of-distribution benchmarks: on TruthfulQA, Qwen2.5-14B achieves plus 17 percent MC1 accuracy (0.500 to 0.585) and plus 49 percent MC2 accuracy (0.357 to 0.531). F-DPO requires no auxiliary reward model, token-level annotations, or multi-stage training.
Similar Papers
DA-DPO: Cost-efficient Difficulty-aware Preference Optimization for Reducing MLLM Hallucinations
Artificial Intelligence
Teaches AI to avoid making up fake answers.
Mask-DPO: Generalizable Fine-grained Factuality Alignment of LLMs
Computation and Language
Makes AI assistants tell the truth more.
Mitigating Hallucination Through Theory-Consistent Symmetric Multimodal Preference Optimization
Artificial Intelligence
Teaches AI to describe pictures without making things up.