Research: Identifying "Ghost Memory" Failure Mode in Named Agents
omarsar0 · x · 2026-07-06
omarsar0 introduces new research revealing the "ghost memory" issue in long-running agents: they confidently recite outdated user facts. This happens because obsolete, current, and transitional records coexist in the memory bank and are retrieved together, misleading the model. The study proposes A-TMA, a state-aware overlay that retains superseded and transitional records instead of deleting them. It builds evidence packages based on the query's state, tagging records as current, historical, or transitional.
Related event: Research Reveals AI Agent 'Ghost Memory' Failure Mode(2 posts)→
More from coding & agent
- Gergely Orosz: Shipping 10x PRs With AI Agents, Sites Fill With Small Regressions — ducha_aiki · 2026-09-11
- Same Echo Maze prompt, three frontier models: all passed visually but shipped the same hidden bug — eyishazyer · 2026-09-11
- Astra storyboards plus Minimax H3 per-shot generation boost video success rates — Hailuo_AI · 2026-09-11
- Codex tip: use Sol with Astra and Luna sub-agents to save usage — pvncher · 2026-09-11
- agents-best-practices: a provider-neutral Agent Skill for designing and auditing agentic harnesses — tom_doerr · 2026-09-11
- Cognition's SWE-2 uses a KKT duality argument in RL to shift the effort Pareto curve — YouJiacheng · 2026-09-11