ActiveSaddler: Automated Curriculum Learning for Agent Harness Optimization

arXiv:2610.00906 · cs.AI, cs.CL, cs.LG, cs.MA, cs.SE · Submitted 2026-10-01 · Read on arXiv

cs.AI, cs.CL, cs.LG, cs.MA, cs.SE

Submitted: 2026-10-01

Updated: 2026-10-01

Code: https://github.com/krafton-ai/kira

Project page: https://aka.ms/ActiveSaddler-website

Terminology

• Accuracy — 23× in this paper · explained in 1 episode
• Trace — 22× in this paper · explained in 2 episodes
• Benchmark — 20× in this paper · explained in 2 episodes
• Generalization — 6× in this paper · explained in 13 episodes
• Robustness — 5× in this paper · explained in 5 episodes
• Ablation — 4× in this paper · explained in 1 episode
• Agent Harness — 4× in this paper · explained in 2 episodes
• Abstraction — 3× in this paper · explained in 1 episode
• Adaptation — 2× in this paper · explained in 1 episode
• Benchmarking — 2× in this paper · explained in 3 episodes
• Generative AI — 2× in this paper · explained in 2 episodes
• Verification — 2× in this paper · explained in 2 episodes
• Verifier — 2× in this paper · explained in 2 episodes
• Ablation Studies — 1× in this paper · explained in 1 episode
• Actor — 1× in this paper · explained in 3 episodes
• Agent Skills — 1× in this paper · explained in 1 episode
• Aggregation — 1× in this paper · explained in 1 episode
• Core — 1× in this paper · explained in 2 episodes
• Distribution Shift — 1× in this paper · explained in 5 episodes
• Failure Taxonomy — 1× in this paper · explained in 2 episodes
• Human-in-the-Loop — 1× in this paper · explained in 3 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes
• Online Adaptation — 1× in this paper · explained in 3 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Reward Hacking — 1× in this paper · explained in 8 episodes
• Self-Evolving Agents — 1× in this paper · explained in 2 episodes

Sources

• Co-Harness: Co-Evolving Harnesses and Model Weights for LLM Agents
• AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence
• Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments
• DREvo: Distilling Recalibrated Historical Experience for Harness Self-Evolution
• SIA: Self Improving AI with Harness & Weight Updates
• Continual Harness: Online Adaptation for Self-Improving Foundation Agents
• WHALE: A Simple Recipe for Joint Harness-Weight Optimization
• Meta-Harness: End-to-End Optimization of Model Harnesses
• Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses
• Teacher-Student Curriculum Learning
• Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces
• TTHE: Test-Time Harness Evolution
• Evolving Agents in the Dark: Retrospective Harness Optimization via Self-Preference
• HARBOR: Automated Harness Optimization
• Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions
• Huxley-G\"odel Machine: Human-Level Coding Agent Development by an Approximation of the Optimal Self-Improving Machine
• Evo-Harness: Context-to-Harness Skill Compilation for Self-Evolving Agents
• Verify Smarter, Evolve Further: Efficient Harness Evolution through Behavior-Aware Verification

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
• VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection