← The AI Systems Map
L0 · LabsInstitution

AI Safety & Interpretability Research (ARC, Redwood, Apollo)

Independent labs probing deception, scheming, interpretability

A cluster of nonprofits and lab teams — the Alignment Research Center, Redwood Research, Apollo Research, and internal interpretability/alignment-science teams — study deceptive alignment, scheming, sandbagging, and mechanistic interpretability that inform frontier deployment decisions. Apollo's 2024-2025 scheming evaluations were cited in model cards. This research directly shapes safety-case arguments.

Focus

Deception / interpretability

Impact

Cited in model cards

How it fits the stack

AI Safety & Interpretability Research (ARC, Redwood, Apollo) with what it depends on (above) and what it feeds (below). The figure renders as a crawlable diagram and upgrades to an interactive 3D graph as it scrolls into view.

partners withAI Safety &InterpretabilityResearch (ARC,LabsAnthropic
AI Safety & Interpretability Research (ARC, Redwood, Apollo)Related

AI Safety & Interpretability Research (ARC, Redwood, Apollo) in the AI stack. AI Safety & Interpretability Research (ARC, Redwood, Apollo) with its immediate upstream dependencies (top) and downstream dependents (bottom) in the AI value chain. Hover a node in 3D, or read the full relationships below.

Graph data (text) — 2 entities, 1 relationships
  • AI Safety & Interpretability Research (ARC, Redwood, Apollo)partners withAnthropic