{"version":1,"type":"story","url":"https://digestai.news/story/qwen-models-struggle-with-outdated-facts-in-retrieved-evidence","json":"https://digestai.news/story/qwen-models-struggle-with-outdated-facts-in-retrieved-evidence.json","markdown":"https://digestai.news/story/qwen-models-struggle-with-outdated-facts-in-retrieved-evidence.md","slug":"qwen-models-struggle-with-outdated-facts-in-retrieved-evidence","headline":"Qwen models struggle with outdated facts in retrieved evidence","summary":"A new study on arXiv finds that AI agents often fail not because they miss relevant facts but because they cannot distinguish outdated information from correct updates. Researchers tested Qwen2.5-7B-Instruct and Qwen3-14B on the FactConsolidation task, where accuracy dropped sharply when evidence sets included competing, older statements—even if the newest fact was present.\n\nWith a 64-token evidence budget, Qwen2.5-7B-Instruct scored 45% accuracy, but performance plummeted to 32% when the budget expanded to 512 tokens. The same issue appeared in a 32K-history test, where BM25-64 outperformed BM25-512. The study suggests that retrieval alone cannot guarantee correct answers if agents cannot filter out stale information.","keyPoints":["Qwen2.5-7B-Instruct accuracy drops from 45% to 32% when evidence budget grows from 64 to 512 tokens","Obsolete facts hurt performance more than unrelated evidence of equal size for Qwen2.5 and Qwen3-14B","Study uses FactConsolidation task from MemoryAgentBench to isolate retrieval vs. fact-staleness issues"],"whyItMatters":"This research highlights a critical flaw in AI agents: retrieving correct facts doesn’t ensure correct answers if outdated versions remain in memory. It could drive better design for fact-checking and temporal reasoning in retrieval-augmented systems.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":[],"models":["Qwen2.5-7B-Instruct","Qwen3-14B"],"people":[]},"firstPublishedAt":"2026-10-28T04:00:00Z","updatedAt":"2026-10-28T04:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"When Retrieval Is Not Enough: Obsolete Facts in Agent Memory","url":"https://arxiv.org/abs/2610.36308","publishedAt":"2026-10-28T04:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Qwen models struggle with outdated facts in retrieved evidence\", 28 October 2026, https://digestai.news/story/qwen-models-struggle-with-outdated-facts-in-retrieved-evidence","publisher":"Digest AI","title":"Qwen models struggle with outdated facts in retrieved evidence","datePublished":"2026-10-28T04:00:00Z","url":"https://digestai.news/story/qwen-models-struggle-with-outdated-facts-in-retrieved-evidence"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}