Brilliant work on using many harness configurations to scale agentic RL.

Many approaches use mainly one fixed harness, which specializes the model but doesn't generalize well. This work uses many harness configurations as a scaling dimension.

But the big question is whether it can generalize to harness configurations never seen in training.

They pick a compact training subset with balanced factorial coverage, and vary one part of the harness at a time on the same task.

Across benchmarks, model sizes, and RL algorithms, this improves results on harness configurations never seen in training. Models trained this way also transfer to unseen external black-box harnesses better than models trained through one complex black-box harness.

If your model will run inside harnesses you did not train it in, treat harness diversity as its own scaling axis.

https://arxiv.org/abs/2610.29458

Chat with Paper: https://academy.dair.ai/papers/harness-scaling-for-generalizable-agentic-reinforcement-learning-2610.29458