Showing cs.AIShow all
2 papers · 1 filter
cs.AI2026
AEGIS: A Backup Reflex for Physical AI
Josef Chen
Long-horizon robot manipulation tends to fail gradually: one bad step degrades the state, and the policy spirals into a basin from which it cannot recover. The failure is often vis…
cs.AI2026
AURA: Action-Gated Memory for Robot Policies at Constant VRAM
Josef Chen
The KV-cache is the right memory for datacenters but the wrong memory for robots. Datacenter inference batches many short requests and resets them, amortizing an attention cache ac…