Researcher flags interpretability paper uses inconsistent S1/S2 vectors across experiments
rgblong · x · 2026-10-09
Researcher rgblong wrapped up an open Q&A discussion with the authors of an interpretability paper, thanking them for the cordial engagement.
In his closing remark, he noted a practical reading difficulty: the paper uses different steering vectors, S1 and S2, for different experiments. While this is clearly flagged, S1 and S2 appear quite different, making it hard to assemble the evidence into one coherent picture.
The thread offers firsthand critical feedback on the paper's experimental design and reflects the field's culture of open engagement.
Related event: Researchers Challenge the 'Pain Axis' Paper as AI Welfare Debate Deepens(19 posts)→
More from Research
- RWTH Aachen's ARROW unifies 3D reconstruction and point tracking from any RGB inputs, sets new SOTA — kwangmoo_yi · 2026-10-09
- HAIPS 2026 workshop lands at COLM tomorrow with top AI privacy researchers — tianshi_li · 2026-10-09
- Tighter Bayes error bounds via generalized Bhattacharyya and Chernoff means — FrnkNlsn · 2026-10-09
- PAMI: part-anchored motion lifts contact recall 14.5% in text-to-HOI generation — _akhaliq · 2026-10-09
- Study finds LLM novelty judges are unstable: scores swing wildly with evaluation design choices — _akhaliq · 2026-10-09
- Google's AMIE lands in The Lancet: 90% match with doctors' final diagnoses in real clinics — sundarpichai · 2026-10-09