Stanford's Scott Linderman joins Engram to build agents with native memory
anshulkundaje · x · 2026-08-18
Stanford professor Scott Linderman is taking a sabbatical to work full-time at Engram. The company published its first research blog, 'Understanding a Law Firm through Study,' showcasing a future where agents are trained with native memory systems. The focus is on building agents that can learn and study using complementary memory architectures.
More from Companies & People
- Analysis suggests OpenAI losses per model, but 2026 agent boom could boost margins to 85% — ben_j_todd · 2026-08-18
- Agent Boom Pushes Frontier Model Gross Margins to Over 85% — ben_j_todd · 2026-08-18
- Bittensor co-founder: building open, permissionless AI you can mine like Bitcoin — markjeffrey · 2026-08-18
- OpenAI funds 14 independent projects on AI, economic opportunity and resilience — OpenAINewsroom · 2026-08-18
- Ship one bet per day: condensing failure into a daily flow — doooyle · 2026-08-18
- KOL hints at potential new move by Qwen team — generativist · 2026-08-18