{"slug": "tev-scaling-long-horizon-agentic-environments-with-dense-progress-rewards", "title": "Tev: Scaling Long-Horizon Agentic Environments with Dense Progress Rewards", "summary": "A 13 October 2026 arXiv paper introduces Tev-Gym, described as the first long-horizon agent training dataset with verifiable, dense, progressive rewards, comprising 1,170 verified task environments across 7 domains. The authors pair Tev-Gym with progress-potential reward shaping (PPRS), which converts a per-step progress metric into a dense reward, and use it to train Tev-27B, a model that matches GPT-6 Astra and Claude Opus 5 at Pass@5 on the Agents' Last Exam computing and math sciences domains (ALE-CM) and surpasses GLM 5.2 and Gemini 3.7 Flash on Terminal-Bench 3.0 and 4.0 subsets. The work targets the sparse-reward and data-scarcity bottlenecks that keep open-source models from handling tasks lasting hours to days.", "body_md": "# Computer Science > Computation and Language\n\n  [Submitted on 13 Oct 2026]\n\n# Title:Tev: Scaling Long-Horizon Agentic Environments with Dense Progress Rewards\n\n[View PDF](http://arxiv.org/pdf/2610.18868v1)\n\n[HTML (experimental)](https://arxiv.org/html/2610.18868v1)\n\nAbstract:Long-horizon agents have drawn growing attention for their ability to work autonomously -- from hours to days -- on economically valuable tasks, a capability essential to automated discovery and recursive self-improvement (RSI). Yet open-source models still struggle on long-horizon tasks, largely for lack of both training data and a working recipe, and scaling both is hard: long, complex tasks are difficult to curate and verify, and their rewards are sparse. We address both. First, we introduce Tev-Gym, the first long-horizon agent training dataset with verifiable, dense, progressive rewards -- 1,170 verified task environments across 7 domains. It is built on a latent-world-first approach that first constructs a latent world: a complete and correct hidden specification of the task domain, workflow, and expected state after each step. Each task is then progressively rendered from this specification into an agent-facing environment, enabling controllable complexity and a natural progress metric. Second, we turn this metric into a training recipe: via progress-potential reward shaping (PPRS), we convert progress into a dense per-step reward that improves training efficiency and makes otherwise-intractable tasks trainable. Training with this recipe yields Tev-27B, which achieves substantial gains on the Agents' Last Exam computing and math sciences domains (ALE-CM), matching GPT-6 Astra and Claude Opus 5 at Pass@5. Despite training only on ALE-related data, it generalizes to the Terminal-Bench (TB) 3.0 and 4.0 subsets, surpassing GLM 5.2 and Gemini 3.7 Flash, and reaches Claude Opus 4.8-level performance on middle-horizon tasks such as TB 2.1.\n    \n\n### Additional Features\n\n### Current browse context:\n\ncs.CL\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/tev-scaling-long-horizon-agentic-environments-with-dense-progress-rewards", "canonical_source": "http://arxiv.org/abs/2610.18868v1", "published_at": "2026-10-14 14:20:21+00:00", "updated_at": "2026-10-15 04:53:18.566176+00:00", "lang": "en", "topics": ["ai-agents", "ai-research", "large-language-models", "machine-learning", "artificial-intelligence"], "entities": ["Tev-Gym", "Tev-27B", "progress-potential reward shaping (PPRS)", "Agents' Last Exam computing and math sciences domains (ALE-CM)", "Terminal-Bench (TB) 3.0 and 4.0", "GPT-6 Astra", "Claude Opus 5", "GLM 5.2"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/tev-scaling-long-horizon-agentic-environments-with-dense-progress-rewards", "markdown": "https://wpnews.pro/news/tev-scaling-long-horizon-agentic-environments-with-dense-progress-rewards.md", "text": "https://wpnews.pro/news/tev-scaling-long-horizon-agentic-environments-with-dense-progress-rewards.txt", "jsonld": "https://wpnews.pro/news/tev-scaling-long-horizon-agentic-environments-with-dense-progress-rewards.jsonld"}}