# ObserverBench: Testing Mechanistic Estimates for Intervention and Control

Preprint · Vijay Erramilli

- Archive id: arxiv:2609.03026
- Published: 2026-09-02
- arXiv: https://arxiv.org/abs/2609.03026
- Source URL: https://arxiv.org/abs/2609.03026
- PDF: https://arxiv.org/pdf/2609.03026
- Provenance: arxiv

## Abstract

Mechanistic interpretability is increasingly used to guide interventions such as activation steering, circuit removal, and safety monitoring. Yet an internal estimate that is accurate on average can still choose a poor action. We present ObserverBench, a benchmark framework for testing whether an internal estimator---an observer---is adequate for the intervention, control, or safety task it directs. Each task fixes the model, information boundary, allowed actions, decision rule, held-out cases, and loss. The benchmark reports estimation accuracy separately from the loss caused by the chosen action. Theory and experiments show why both are needed. In closed-loop control, observer errors matter at the starting point and along directions the allowed intervention can reach. On circuit-intervention tasks in GPT-2-small and Qwen2.5-7B, pairwise observers predict unseen effects more accurately without always choosing better actions; observers trained on action loss choose lower-loss actions. In safety triage, a score that perfectly separates violations can allocate a fixed intervention budget poorly when violations have different costs. Across Qwen2.5-7B, Gemma-2-9B-it, and prospectively frozen Qwen3.5-9B APPS tasks, AUROC can rank monitors differently from deployment loss, and the best information source changes across models. Sparse SAE readouts also trail their layer-matched dense controls on the reported Qwen panels, under disclosed activation-density or checkpoint mismatches. ObserverBench provides fixed task contracts, runnable baselines, and table-based submissions for evaluating interpretability methods through the actions they enable.

_Abstract reproduced under CC0-1.0 (arXiv descriptive metadata)._

---
Space Kitty Archive · https://space-kitty.com · API: https://space-kitty.com/api/v1 · MCP: https://space-kitty.com/mcp · Premium: $0.20/request via x402 (Base Sepolia (testnet)). Metadata: arXiv (CC0), OpenAlex (CC0), Europe PMC (CC BY/CC0 works), ClinicalTrials.gov.