e8a500138d
- Source: inbox/queue/2026-04-09-lindsey-representation-geometry-alignment-probing.md - Domain: ai-alignment - Claims: 2, Entities: 0 - Enrichments: 0 - Extracted by: pipeline ingest (OpenRouter anthropic/claude-sonnet-4.5) Pentagon-Agent: Theseus <PIPELINE>
17 lines
1.9 KiB
Markdown
17 lines
1.9 KiB
Markdown
---
|
|
type: claim
|
|
domain: ai-alignment
|
|
description: The read-only advantage of geometric probing is structurally coupled to infrastructure control, bounding where the method applies
|
|
confidence: experimental
|
|
source: "Lindsey & Garriga-Alonso (Anthropic), arxiv 2604.02891"
|
|
created: 2026-04-09
|
|
title: Trajectory geometry probing requires white-box access to all intermediate activations, making it deployable in controlled evaluation contexts but not in adversarial external audit scenarios
|
|
agent: theseus
|
|
scope: structural
|
|
sourcer: Jack Lindsey, Adria Garriga-Alonso (Anthropic)
|
|
related_claims: ["[[pre-deployment-AI-evaluations-do-not-predict-real-world-risk-creating-institutional-governance-built-on-unreliable-foundations]]"]
|
|
---
|
|
|
|
# Trajectory geometry probing requires white-box access to all intermediate activations, making it deployable in controlled evaluation contexts but not in adversarial external audit scenarios
|
|
|
|
Geometric probing requires white-box access to all intermediate activations across reasoning steps — it cannot be deployed as a behavioral API test. This requires inference-time activation recording infrastructure, which means the method works in evaluation contexts where the model operator controls the infrastructure but fails in adversarial external audit contexts where auditors lack internal access. The read-only property that prevents adversarial targeting is therefore structurally coupled to infrastructure control: defenders who can monitor trajectory geometry are necessarily the same parties who control the deployment infrastructure. This creates a fundamental boundary condition: trajectory geometry is a tool for internal alignment evaluation and monitoring, not for external oversight or third-party auditing. The method provides a partial escape from the SAE dual-use trap, but only within the trust boundary of organizations that already control model deployment.
|