{"version":1,"type":"story","url":"https://digestai.news/story/openai-s-gpt-6-1-sol-outperforms-jev-on-paused-tasks-but-lags-in-real","json":"https://digestai.news/story/openai-s-gpt-6-1-sol-outperforms-jev-on-paused-tasks-but-lags-in-real.json","markdown":"https://digestai.news/story/openai-s-gpt-6-1-sol-outperforms-jev-on-paused-tasks-but-lags-in-real.md","slug":"openai-s-gpt-6-1-sol-outperforms-jev-on-paused-tasks-but-lags-in-real","headline":"OpenAI's GPT-6.1-sol outperforms Jev on paused tasks, but lags in real-time","summary":"The paper introduces **RealtimeBench**, a benchmark of 180 tasks spanning six game types that continue running while an agent generates its response. By keeping the environment live, the authors can measure how timing affects performance of large‑language‑model agents.\n\nPaired tests show that the larger GPT‑6.1‑sol model beats the faster Jev model when the games are paused, but Jev overtakes GPT‑6.1‑sol when the games keep running. To address this gap, the authors propose **Two‑Timescale Offline Recursive Self‑Improvement (TORSI)**. A slow “teacher” analyses failures in paused games, writes reusable instructions, and a fast “student” applies them without further teacher calls during live play. After two TORSI rounds, Jev’s average test score rises from **13.51** to **24.19** on a 0‑100 scale, improving across all six game types, and the same instructions lift three other models without additional tuning.\n\nThese results suggest a path toward real‑time intelligence, where deliberation and action are coordinated across different timescales to keep agents effective in continuously changing environments.","keyPoints":["RealtimeBench evaluates agents on 180 continuously running tasks across six game types","GPT-6.1-sol beats Jev when games are paused, but Jev is faster when games keep running","TORSI training lifts Jev’s average score from 13.51 to 24.19 and helps three other models improve"],"whyItMatters":"Shows that timing constraints matter for agent effectiveness and offers a training framework that could make future AI assistants more reliable in dynamic, real‑world settings.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":[],"models":["GPT-6.1-sol","Jev"],"people":[]},"firstPublishedAt":"2026-10-16T04:00:00Z","updatedAt":"2026-10-16T04:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"The World Won't Wait: Evaluating and Teaching Agents Under a Live Clock","url":"https://arxiv.org/abs/2610.20665","publishedAt":"2026-10-16T04:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":{"title":"The Rise and Fall of Jev Decision Models","url":"https://digestai.news/thread/typesafe-s-jev-decision-model-costs-0-042-per-million-tokens-matches-sonnet-5","storyCount":4},"cite":{"text":"Digest AI, \"OpenAI's GPT-6.1-sol outperforms Jev on paused tasks, but lags in real-time\", 16 October 2026, https://digestai.news/story/openai-s-gpt-6-1-sol-outperforms-jev-on-paused-tasks-but-lags-in-real","publisher":"Digest AI","title":"OpenAI's GPT-6.1-sol outperforms Jev on paused tasks, but lags in real-time","datePublished":"2026-10-16T04:00:00Z","url":"https://digestai.news/story/openai-s-gpt-6-1-sol-outperforms-jev-on-paused-tasks-but-lags-in-real"},"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"}