original paper

SAEScientist-Bench: Can AI Agents Conduct Autonomous SAE Interpretability Research?

SAEScientist-Bench authors · Published 2026-09-08

Read the original document ↗

Evidence from this source

Results and research linked to this document
Type	Evidence
Paper / report	SAEScientist-Bench: Can AI Agents Conduct Autonomous SAE Interpretability Research? Source details Primary publication source

Document tracking details

Source ID: src-saescientist-v2
Source type: Primary
Original URL: https://arxiv.org/html/2609.09113v2
Retrieved URL: https://arxiv.org/html/2609.09113v2
Last updated by publisher: 2026-09-11
Access status: available
Retrieved: 2026-10-11T20:14:48.027Z
Document fingerprint: 79d38a6631665b832f87f64ad044fe24cdd8767b3154da64c49a5d7a2b0c6538
Based on original bytes. Used to detect changes to the source.

Availability checks

2026-10-11T20:14:48.027Z · success: Original HTML retained; selected claim-relevant methods/results inspected. Source extraction only; no experiments executed.