aiqbench analysis
TaskPGM introduces a probabilistic graphical model that optimizes supervised fine-tuning budgets by explicitly modeling inter-task relationships rather than relying on simple heuristics. By leveraging behavioral divergences within a Markov random field, the framework effectively balances task coverage against redundancy, yielding measurable performance gains across multiple model families. This approach offers a theoretically grounded method for improving transfer learning efficiency and reducing wasted computational resources on redundant training data.
Learning Task Mixtures from Task Affinities: A Probabilistic Graphical Model for Supervised Fine-Tuning
arXiv:2507.12612v4 Announce Type: replace Abstract: Supervised fine-tuning performance for large language models depends strongly on how training budget is distributed across a heterogeneous set of tasks. In practice, mixtures are often fixed using simple heuristics (e.g., uniform or size-proportional sampling) that ignore task interactions, which can hurt transfer and waste budget on redundant sources. We introduce TaskPGM, a framework for learning continuous task mixtures via an energy-based model over tasks. Tasks form the nodes of a Markov random field: unary potentials capture per-task utility, and pairwise potentials encode inter-task relationships using behavioral divergences computed from predictive distributions of single-task fine-tuned models (e.g., Jensen--Shannon divergence and pointwise mutual information). Optimizing this objective yields mixtures that balance coverage against redundancy. We show that the resulting set function is weakly submodular under budget constraints, enabling approximation guarantees for discrete selection variants. Across multiple model families (LLaMA-7B, Qwen2-7B) and evaluation suites (BIG-Bench Hard), TaskPGM improves over standard mixing strategies and provides interpretable structure over task interactions.
Abstract:Supervised fine-tuning performance for large language models depends strongly on how training budget is distributed across a heterogeneous set of tasks. In practice, mixtures are often fixed using simple heuristics (e.g., uniform or size-proportional sampling) that ignore task interactions, which can hurt transfer and waste budget on redundant sources. We introduce TaskPGM, a framework for learning continuous task mixtures via an energy-based model over tasks. Tasks form the nodes of a Markov random field: unary potentials capture per-task utility, and pairwise potentials encode inter-task relationships using behavioral divergences computed from predictive distributions of single-task fine-tuned models (e.g., Jensen--Shannon divergence and pointwise mutual information). Optimizing this objective yields mixtures that balance coverage against redundancy. We show that the resulting set function is weakly submodular under budget constraints, enabling approximation guarantees for discrete selection variants. Across multiple model families (LLaMA-7B, Qwen2-7B) and evaluation suites (BIG-Bench Hard), TaskPGM improves over standard mixing strategies and provides interpretable structure over task interactions.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
I.2.7; I.2.6; I.2.4
arXiv:2507.12612 [cs.LG]
(or arXiv:2507.12612v4 [cs.LG] for this version)
Full text under the publisher's open licence · 174 words
Related coverage
Learning Task Mixtures from Task Affinities: A Probabilistic Graphical Model for Supervised Fine-Tuning
arXiv · Sep 30, 2026
Exploding variance of means of exponentials: least-squares to the rescue – Machine Learning Research Blog
francisbach.com · Sep 27, 2026
Salesforce after Dreamforce: How the CRM giant can grow beyond its own interface
SiliconANGLE · Sep 19, 2026
At SIGGRAPH, NVIDIA Advances Graphics and Simulation With Agentic and Physical AI
NVIDIA · Jul 20, 2026