Harness Evolution Hits a Ceiling: When Weight Training Should Begin

arXiv:2610.11655 · cs.AI, cs.CL, cs.LG · Submitted 2026-10-08 · Read on arXiv

cs.AI, cs.CL, cs.LG

Submitted: 2026-10-08

Updated: 2026-10-08

Comments: 21 pages, 7 figures, 15 tables

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Ablation — 17× in this paper · explained in 1 episode
• Benchmark — 16× in this paper · explained in 2 episodes
• Fine-Tuning — 12× in this paper · explained in 5 episodes
• Accuracy — 8× in this paper · explained in 1 episode
• LoRA Adapters — 6× in this paper · explained in 3 episodes
• Trace — 3× in this paper · explained in 2 episodes
• Benchmarking — 2× in this paper · explained in 4 episodes
• Credit assignment — 2× in this paper · explained in 3 episodes
• Frozen model — 2× in this paper · explained in 3 episodes
• Adaptation — 1× in this paper · explained in 1 episode
• Agentic System — 1× in this paper · explained in 2 episodes
• Distillation — 1× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Failure Taxonomy — 1× in this paper · explained in 3 episodes
• LLM judge — 1× in this paper · explained in 2 episodes
• Multi-Agent System — 1× in this paper · explained in 7 episodes
• Provenance — 1× in this paper · explained in 5 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Transformer — 1× in this paper · explained in 2 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• Why Do Multi-Agent LLM Systems Fail?
• FireAct: Toward Language Agent Fine-tuning
• Co-Harness: Co-Evolving Harnesses and Model Weights for LLM Agents
• TRAIL: Trace Reasoning and Agentic Issue Localization
• QLoRA: Efficient Finetuning of Quantized LLMs
• LoRA: Low-Rank Adaptation of Large Language Models
• Automated Design of Agentic Systems
• AI Agents That Matter
• WHALE: A Simple Recipe for Joint Harness-Weight Optimization
• The Interplay of Harness Design and Post-Training in LLM Agents
• Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses
• AgentBench: Evaluating LLMs as Agents
• AgentBoard: An Analytical Evaluation Board of Multi-turn LLM Agents
• Adding Error Bars to Evals: A Statistical Approach to Language Model Evaluations
• MemGPT: Towards LLMs as Operating Systems
• WebRL: Training LLM Web Agents via Self-Evolving Online Curriculum Reinforcement Learning
• Locating Hidden Failures Makes Long-Horizon Agents More Reliable
• Proximal Policy Optimization Algorithms
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• Hindsight Credit Assignment for Long-Horizon LLM Agents

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
• Agents' Last Exam
• 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