ParetoSaddler: Pareto-Guided Optimization of Agent Harnesses for Quality and Cost
cs.AI, cs.CL, cs.LG, cs.MA
Submitted: 2026-11-01
Updated: 2026-11-01
Code: https://github.com/openclaw/openclaw
Project page: https://aka.ms/ParetoSaddler-website
Terminology
Sources
- EvoSkill: Automated Skill Discovery for Multi-Agent Systems
- MO-CAPO: Multi-Objective Cost-Aware Prompt Optimization
- SkillMOO: Multi-Objective Optimization of Agent Skills for Software Engineering
- HarnessEvolve: Learning from Reference Trajectories for Reliable Agent Self-Evolution
- Meta-Harness: End-to-End Optimization of Model Harnesses
- AutoHarness: improving LLM agents by automatically synthesizing a code harness
- ClawMark: A Living-World Benchmark for Multi-Turn, Multi-Day, Multimodal Coworker Agents
- Trace2Skill: Distill Trajectory-Local Lessons into Transferable Agent Skills
- Evolving Agents in the Dark: Retrospective Harness Optimization via Self-Preference
- SkillOpt: Executive Strategy for Self-Evolving Agent Skills
- TextGrad: Automatic "Differentiation" via Text
- Hyperagents
- Externalization in LLM Agents: A Unified Review of Memory, Skills, Protocols and Harness Engineering
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- 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
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents