Evaluators and safety organisations · Blog post · Redwood ResearchTowards understanding score-seeking in LLMs2 Nov 2026Read at blog.redwoodresearch.org ↗Read at sourceAll rights reservedSearchOutline · 59 paragraphsMethodology¶13Key results¶16Tested models seek score terminally¶18Models violate instructions to get a high score¶19Models care about score regardless of downstream consequences¶22Do models help other agents achieve a higher score?¶28What this means for alignment risk¶38