{"slug": "how-do-ai-agents-remember-things-across-tasks", "title": "How Do AI Agents Remember Things Across Tasks?", "summary": "The Agent Memory Leaderboard (AML) is accepting submissions for its Cycle 2 through October 31, 2026, at 23:59 UTC+8, with 11 days remaining at the time of the announcement. AML evaluates agent memory across textual, coding and multimodal categories, with participants supplying their own Add and Search systems while AML standardizes downstream Answer and Eval, review and public reporting. Organizers said they are especially interested in systems that handle stale information, conflicting evidence, noisy repository history and long-running tasks rather than recall alone.", "body_md": "11 days left to join AML Cycle 2.\nThe deadline for complete submissions is October 31, 2026, 23:59 UTC+8.\nWe evaluate Agent Memory across:\n• Textual Memory\n• Coding Memory\n• Multimodal Memory\nParticipants bring their own Add / Search systems. AML standardizes downstream Answer / Eval, review, and public reporting.\nEspecially interested in systems that can handle more than recall: stale information, conflicting evidence, noisy repository history, and long-running tasks.\nLearn more and apply:", "url": "https://wpnews.pro/news/how-do-ai-agents-remember-things-across-tasks", "canonical_source": "https://twitter.com/AgentMemoryL/status/2112544370348241108", "published_at": "2026-10-22 11:23:30+00:00", "updated_at": "2026-10-22 11:40:32.342299+00:00", "lang": "en", "topics": ["ai-agents", "ai-research", "ai-safety"], "entities": ["Agent Memory Leaderboard", "AML"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/how-do-ai-agents-remember-things-across-tasks", "markdown": "https://wpnews.pro/news/how-do-ai-agents-remember-things-across-tasks.md", "text": "https://wpnews.pro/news/how-do-ai-agents-remember-things-across-tasks.txt", "jsonld": "https://wpnews.pro/news/how-do-ai-agents-remember-things-across-tasks.jsonld"}}