Weaker AI models improve task success with expert-guided corrections
Co-Evolving Harnesses and Models: On-Policy Correction Helps Weaker Models Catch Up Where Imitation Fails
Artificial Intelligence
Summary
When smaller AI models try to do complex tasks, they rely on surrounding tools and instructions called harnesses to help them succeed. The authors found that just copying how a stronger AI model plans often makes the weaker model perform worse because the weaker model cannot do everything the stronger one does. To fix this, they created a method where the weaker model runs its own plan but gets help only on parts where it struggles, using corrections from the stronger model. This way, smaller models can learn better and work well with their own harness setup.
What this means in practice
- •For enterprise ai development teams: Improve the performance of smaller AI models on business-specific tasks by co-evolving task scaffolding and model fine-tuning with expert-guided corrections.
- •For customer support automation teams: Enable cost-effective deployment of AI assistants that combine lightweight models with tailored execution setups and corrected guidance for reliable task handling.
Authors
Zhou Yu, Bin Bi, Shiva Kumar Pentyala, Shubham Mehrotra, Sougata Chaudhuri, Shilpa Bhagavath, Zeyuan Chen, Ran Xu, Phil Mui, James Zhu, Sitaram Asur
Abstract
Agent harnesses (the system prompt, tool set, execution hooks, and context-management scaffolding around a model) are a critical determinant of agentic task success. Automated harness evolution can enable smaller models to perform well on domain-specific tasks at a fraction of frontier-model cost. Since both the harness and model weights shape behavior, we ask how harness evolution and lightweight fine-tuning should be combined. Across seven enterprise agent tasks, we first evolve a harness with the weaker model, then find that a stronger expert often uses it more effectively, suggesting expert supervision could close the remaining gap. However, training the weaker model on the expert's complete trajectories under the evolved harness backfires: performance regresses on all seven tasks by 4 to 30 points across Qwen3-Coder and Gemma 4, even though the same procedure helps under the unevolved harness. Our analysis shows that imitation transfers knowledge and increases scaffold usage, but disrupts model-harness fit: the weaker model adopts the expert's planning strategy without the competence to execute it and no longer matches the harness evolved around its native planning style. We therefore develop an on-policy expert-correction pipeline, automated by a meta-level MLE agent, that localizes the failing turn in the weaker model's own rollout and asks the expert to rewrite only that turn. This preserves the model's planning style and combines the gains of harness evolution and model adaptation. Our results identify and resolve a source of contention between harness and weight updates, yielding a compatibility-preserving recipe for economical co-evolution on domain-specific enterprise tasks.