Meta proposes AI Research Preference Models to help agents select high-value experiments
omarsar0 · x · 2026-09-02
Meta discusses a key bottleneck for long-horizon research agents: selecting which experiments to run given limited GPU budget. The paper introduces "AI Research Preference Models" trained to predict the most promising candidate solutions before execution. Two variants are proposed: an inference-only model that reasons over plans and history, and an agentic model that runs small-scale pilot experiments first. Integrated into AIRA-dojo and measured on AIRS-Bench, the approach improved the average normalized score from 0.684 to 0.711 and 0.729, outperforming unguided agents.
More from coding & agent
- Andrew Ng: Master software engineering fundamentals to steer AI agents effectively — DeepLearningAI · 2026-09-02
- GitHub CLI adds --attach flag for media uploads in issues and PRs — mariorod1 · 2026-09-02
- Replit MCP Launches: Control Powerful Agents from Anywhere — amasad · 2026-09-02
- Claude Code 2.1.258 Released with macOS 12 Fixes — ClaudeCodeLog · 2026-09-02
- Claude Code 2.1.258 Released, Fixes macOS 12 Launch Bug — ClaudeCodeLog · 2026-09-02
- Automating Boring Business Tasks with 5 Skydive Agents — nima_owji · 2026-09-02