GPT-6 Astra reportedly leaks unfiltered neuralese traces

Users including Ivan Bezdomny observed that Google's Astra model appears to leak its internal neuralese thought traces in plain output, with sample internal notes surfacing unfiltered, raising concerns about internal reasoning exposure.

2026-09-06 ~ 2026-09-06 · 2 related posts

• GPT-6 Astra appears to leak neuralese internal thinking traces openly — ivan_bezdomny · 2026-09-06
• Astra appears to leak raw neuralese thinking traces in internal notes — ivan_bezdomny · 2026-09-06