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

    Argon is said to be the second model family to officially, effectively max out ARC-AGI-3 using a provider adapter

    A post calls GDM’s results impressive and notes a “backwards bend” in action efficiency.

    François CholletFC
    👩‍💻 Paige Bailey👩‍💻P
    Mike KnoopMK
    8 Sources, 2h ago, first seen 2h ago

    TLDR

    A post says Argon is just the second model family to officially, effectively max out ARC-AGI-3 using a provider adapter. It praises GDM’s results and says the same “backwards bend” is appearing in action efficiency.

    Combined views

    25.1K

    8 Sources, first seen 2h ago

    514 likes34 comments38 saves30 reposts

    Combined views

    25.1K

    8 Sources, first seen 2h ago

    514 likes34 comments38 saves30 reposts
    Image: @sundeep on X

    Sentiment

    Positive22.4%77.6%Negative

    Summary

    Positive accounts welcomed Argon matching Astra on ARC-AGI-3 as proof the frontier race is tightening, while negative replies called the benchmark useless or unrelated to real progress.

    Based on 32 sentiment-bearing replies from 30 accounts across 3 conversations.

    Featured Source

    Sentiment

    Positive22.4%77.6%Negative

    Summary

    Positive accounts welcomed Argon matching Astra on ARC-AGI-3 as proof the frontier race is tightening, while negative replies called the benchmark useless or unrelated to real progress.

    Based on 32 sentiment-bearing replies from 30 accounts across 3 conversations.

    8 Sources

    Mike Knoop@mikeknoopArgon is just the second model family to officially, effectively, max out ARC-AGI-3 using a provider adapter. Impressive results from GDM here! Also notable that we're seeing the same "backwards bend" on action efficiency.2h
    Lisan al Gaib@scaling01I'm not quite sure what role loopers played in saturating ARC-AGI-3 but I definitely believe it's solvable by just a shit ton of RL on games. loopers probably just made it a lot cheaper overall I think I am at 85% RL on games and 15% looping for what caused this sudden jump, mostly because of the Opus 5 datapoint and general trends in AI (everything can be RL'd)2h
    François Chollet@fcholletArgon has a similar performance profile on ARC-AGI-3 as Astra. Google is back at the frontier.1h
    Melvin Johnson@melvinjohnsonpStrong showing from Gemini 4 Argon on ARC-AGI-3!1h
    👩‍💻 Paige Bailey@DynamicWebPaige🙌 @GoogleDeepMind is the second frontier AI lab to score above 95% on ARC-AGI-3, with Gemini 4 Argon.51m
    Haydn Belfield@HaydnBelfieldCool finding51m
    sunny madra@sundeepGetting close to the promise land:43m
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI

    8 Sources

    Mike Knoop@mikeknoopArgon is just the second model family to officially, effectively, max out ARC-AGI-3 using a provider adapter. Impressive results from GDM here! Also notable that we're seeing the same "backwards bend" on action efficiency.2h
    Lisan al Gaib@scaling01I'm not quite sure what role loopers played in saturating ARC-AGI-3 but I definitely believe it's solvable by just a shit ton of RL on games. loopers probably just made it a lot cheaper overall I think I am at 85% RL on games and 15% looping for what caused this sudden jump, mostly because of the Opus 5 datapoint and general trends in AI (everything can be RL'd)2h
    François Chollet@fcholletArgon has a similar performance profile on ARC-AGI-3 as Astra. Google is back at the frontier.1h
    Melvin Johnson@melvinjohnsonpStrong showing from Gemini 4 Argon on ARC-AGI-3!1h
    👩‍💻 Paige Bailey@DynamicWebPaige🙌 @GoogleDeepMind is the second frontier AI lab to score above 95% on ARC-AGI-3, with Gemini 4 Argon.51m
    Haydn Belfield@HaydnBelfieldCool finding51m
    sunny madra@sundeepGetting close to the promise land:43m
    Today's Rank

    #5

    Today's Rank

    #5