Training the Whole Agent, Not Just the Model
We introduce HarnessAdapt-RL, which trains the policy and adapts its harness during reinforcement learning.
ProofPaper ↗
Key points
- Training coding agents typically updates the policy while keeping the harness fixed.
- During training, trajectories reveal flawed repair strategies and interaction problems that vary across tasks and change as the policy learns.
- For interaction problems recurring across tasks, a separate teacher model evolves tools and hooks, keeping each change only if it improves the observed pass rate or the observed trainable rate without lowering either.
- With a 9B base model, HarnessAdapt-RL improves pass@1 over GRPO by 9.8, 12.6, and 6.4 percentage points on SWE-bench Verified, Multilingual, and Pro, and outperforms GRPO with the human-designed OpenHands harness.
Sources (1)
- [1]Training the Whole Agent, Not Just the ModelHugging Face Daily Papers · Oct 20, 12:00 AM
We introduce HarnessAdapt-RL, which trains the policy and adapts its harness during reinforcement learning.
Training coding agents typically updates the policy while keeping the harness fixed.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 19, 2026HarnessRL: Co-Evolving Policies with a Just-in-Time Training Harness for Open-Ended RL
- Oct 19, 2026Which Agent's Skill Should Change? Group-Relative Credit Assignment for Evolving Multi-Agent Teams
- Oct 18, 2026Bud ContextCompact 2.0: Context Compaction for Agent Success and Accuracy with Small Language Models and Constrained Context Windows
- Oct 18, 2026EG-GRPO: Evidence-Guided Step-level Credit Assignment for Deep Search Agents
- Oct 18, 2026Trackable Skill Library Lifecycle Management for Agentic Reinforcement Learning
- Oct 18, 2026StopLoss: Teaching Agents to Predict Failure and Stop Early