Microsoft's ParetoSaddler cuts agent cost 42% while lifting GAIA2 quality to 62.89%
omarsar0 · x · 2026-11-04
Elvis omarsar0 breaks down Microsoft's ParetoSaddler: apply cost preference only at the end of harness-optimization search, since per-step cost filtering discards harnesses that lead to better ones. Each iteration edits a parent harness, runs both versions repeatedly, and uses Bayesian bootstrap Pareto support to promote candidates; deployment choice is separate. Results: GAIA2 quality 56.22%→62.89% with cost $0.78→$0.45 per task; ClawMark 70.86%→81.35% with cost $1.17→$1.01. Pareto-selected harnesses beat quality-only optimization on both axes. Code to be released.
More from coding & agent
- Free no-sign-in MCP server compiles LaTeX (pdflatex/xelatex) and explains errors in plain words — PurplelinkPL · 2026-11-04
- Multi-agent review pipeline: Claude coordinates, Grok reviews, Codex fixes — PawelHuryn · 2026-11-04
- WoW x League of Legends Arena mashup: roguelike 2v2v2v2 PvP with per-round drafts — TAbrodi · 2026-11-04
- Kimi K3.1 Pro weights drop tomorrow: 2.8T params, 1M-token context — togethercompute · 2026-11-04
- 86% of orgs run AI agents, but under 18% can automatically stop a misbehaving one — amasad · 2026-11-04
- CMU Study: Harness Engineering Boosts Scientific Agents Without Retraining — niloofar_mire · 2026-11-04