‹ All stories

IQuest-Q1

Model · 1 update · last updated Sep 29

Where it stands

As of Sep 29, IQuest-Q1 is a 320B-parameter sparse-MoE model (about 15B active) reported by QbitAI, with demos spanning game generation, three-body equation solving and RL training-data debugging.

Timeline · newest first
  1. Sep 29
    Reported by QbitAI

    IQuest-Q1: a 320B MoE model (15B active) that spots RL data bugs and builds games from prompts, per QbitAI

    QbitAI reports IQuest-Q1, a decoder-only sparse-MoE model with about 320B total parameters and roughly 15B active. Demonstrations include generating a playable racing game from a single prompt, solving 100k-step three-body equations, and spotting a hidden-space bug in RL training data. QbitAI says it performs well on benchmarks including NL2Repo, CyberGym, Terminal-Bench 2.1, DeepSWE v1.1 and JobBench.

    Source: QbitAI
How we label
Official
An institution announces something that has already happened, through its own channel or a founder or executive account.
Statement by …
A named person's views, predictions, plans or comments, including remarks reported by media when the original is unavailable.
Reported by …
Only media or another third party is saying it. We name the outlet.
Unconfirmed
Circulating online (posts, comments, leaks). No company, named person or outlet has confirmed it.

Items older than 72 hours never go in Top stories and show their original date.

DailyCitedRSSAboutPrivacy