InfiFPO: Implicit Model Fusion via Preference Optimization in Large Language Models
read the original abstract
Model fusion combines multiple Large Language Models (LLMs) with different strengths into a more powerful, integrated model through lightweight training methods. Existing works on model fusion focus primarily on supervised fine-tuning (SFT), leaving preference alignment (PA) --a critical phase for enhancing LLM performance--largely unexplored. The current few fusion methods on PA phase, like WRPO, simplify the process by utilizing only response outputs from source models while discarding their probability information. To address this limitation, we propose InfiFPO, a preference optimization method for implicit model fusion. InfiFPO replaces the reference model in Direct Preference Optimization (DPO) with a fused source model that synthesizes multi-source probabilities at the sequence level, circumventing complex vocabulary alignment challenges in previous works and meanwhile maintaining the probability information. By introducing probability clipping and max-margin fusion strategies, InfiFPO enables the pivot model to align with human preferences while effectively distilling knowledge from source models. Comprehensive experiments on 11 widely-used benchmarks demonstrate that InfiFPO consistently outperforms existing model fusion and preference optimization methods. When using Phi-4 as the pivot model, InfiFPO improve its average performance from 79.95 to 83.33 on 11 benchmarks, significantly improving its capabilities in mathematics, coding, and reasoning tasks.
This paper has not been read by Pith yet.
Forward citations
Cited by 5 Pith papers
-
Discovering Physical Directions in Weight Space: Composing Neural PDE Experts
Fine-tuning neural PDE operators to regime endpoints reveals a physical direction in weight space that CCM uses to compose accurate merged models for new or extrapolated regimes from metadata or short prefixes.
-
FeatCal: Feature Calibration for Post-Merging Models
FeatCal reduces feature drift in merged models via layer-wise closed-form calibration on a small dataset, outperforming prior post-merging methods on CLIP and GLUE benchmarks with high sample efficiency.
-
E-PMQ: Expert-Guided Post-Merge Quantization with Merged-Weight Anchoring
E-PMQ improves 4-bit quantization accuracy on merged models by 8-42 points across CLIP and GLUE tasks through expert-guided calibration and merged-weight anchoring.
-
Geometry Conflict: Explaining and Controlling Forgetting in LLM Continual Post-Training
Forgetting in LLM continual post-training is a geometry conflict between task-induced covariance structures and the evolving model state, controlled by gating Wasserstein barycenter merging on measured conflict.
-
Model Merging Scaling Laws in Large Language Models
Empirical scaling laws for LLM merging show a size-dependent floor and 1/k-like tail in cross-entropy loss that holds across architectures and merging methods.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.