• Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    AI
    Announcement

    GPT-6 Astra reportedly collects 61 MazeBench gems with WorldBox, versus 23 without

    EnactraAI says its spatial-reasoning sandbox helps agents manage long-term memory, infer mechanics and simulate reasoning.

    Lianhui QinLQ
    6 Sources, 2h ago, first seen 2h ago

    TLDR

    EnactraAI says GPT-6 Astra collected 61 gems with WorldBox, compared with 23 without it, on MazeBench. The company describes MazeBench as a 256-room test of long-horizon spatial reasoning and WorldBox as a sandbox that helps agents manage memory, infer mechanics and simulate reasoning.

    Combined views

    162

    6 Sources, first seen 2h ago

    2 likes1 saves1 reposts

    Combined views

    162

    6 Sources, first seen 2h ago

    2 likes1 saves1 reposts
    Image: @EnactraAI on X

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Featured Source

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    6 Sources

    Enactra AI@EnactraAI🚨New level unlocked! 🚨 We doubled GPT‑6 Astra’s score on MazeBench with WorldBox, our agentic sandbox for spatial reasoning. Astra: 23 gems 💎💎 Astra+WorldBox: 61 gems 💎💎💎💎💎 MazeBench, by @patience_cave, is a tough test of long-horizon spatial reasoning: 256 interconnected rooms where even Opus 5.5 used 2.2B tokens and collected just 13 of 100 gems. WorldBox provides agents with an agentic sandbox, for managing long-term memory, inferring complex mechanics, and simulating reasoning. Watch the sandbox take shape and the agent tackle increasingly tricky puzzles. 🧵2h
    Zhiting Hu@ZhitingHuRT @EnactraAI: 🚨New level unlocked! 🚨 We doubled GPT‑6 Astra’s score on MazeBench with WorldBox, our agentic sandbox for spatial reasoning.…1h
    Lianhui Qin@LianhuiqLong-horizon spatial reasoning remains a major challenge, even for the strongest frontier AI models. Impressive results from @EnactraAI! 🚀 WorldBox boosts GPT-6 Astra from 23 to 61 gems on MazeBench, a 2.7× improvement! Exciting to see how memory, world modeling, and simulation can help AI agents tackle much harder problems.40m
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI

    6 Sources

    Enactra AI@EnactraAI🚨New level unlocked! 🚨 We doubled GPT‑6 Astra’s score on MazeBench with WorldBox, our agentic sandbox for spatial reasoning. Astra: 23 gems 💎💎 Astra+WorldBox: 61 gems 💎💎💎💎💎 MazeBench, by @patience_cave, is a tough test of long-horizon spatial reasoning: 256 interconnected rooms where even Opus 5.5 used 2.2B tokens and collected just 13 of 100 gems. WorldBox provides agents with an agentic sandbox, for managing long-term memory, inferring complex mechanics, and simulating reasoning. Watch the sandbox take shape and the agent tackle increasingly tricky puzzles. 🧵2h
    Zhiting Hu@ZhitingHuRT @EnactraAI: 🚨New level unlocked! 🚨 We doubled GPT‑6 Astra’s score on MazeBench with WorldBox, our agentic sandbox for spatial reasoning.…1h
    Lianhui Qin@LianhuiqLong-horizon spatial reasoning remains a major challenge, even for the strongest frontier AI models. Impressive results from @EnactraAI! 🚀 WorldBox boosts GPT-6 Astra from 23 to 61 gems on MazeBench, a 2.7× improvement! Exciting to see how memory, world modeling, and simulation can help AI agents tackle much harder problems.40m
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet