statement
Pre-training is complete and 4.8 is now in the RL stage, with evaluation scores starting to rise.
A statement by 马斯克. Statements are as reported by the cited publishers — not verified by DecisionGateway.
预训练已完成,4.8 现进入 RL 阶段,评测分数开始上升。
Discussion
Related claims
4.8 is not yet impressive, but the foundation is right.马斯克 · statementThe post's author concludes that in supervised learning a distribution shift between train and test is usually a problem, while for RL training it is part of the solution.כותב הפוסט · statementGrok 4.8 has entered the reinforcement learning stage.马斯克 · statementOn algorithmic tasks (CLRS-Text), ten such layers raised an 86M-parameter model from 55% to 84%, after regular training got stuck at 55% and even training twice as long did not move it.post author · statementThe author has high confidence the competitive bar just moved and medium confidence Anthropic's prerelease results translate cleanly at scale.post author · predictionThe models trained on the easier examples are tested on problems harder than those they saw in post-training and still succeed much more than models trained directly on the test difficulty level, without receiving full solutions or feedback on intermediate steps.המחברים · investigation