AI/ ai · machine-learning · self-play · language-models

A Random Word Mask Stops AI Self-Play From Repeating Itself

Researchers trained Qwen3 models with vocabulary dropout, a shifting output mask, and measured a 4.4-point average gain on math benchmarks.

Randomly masking which words a language model can use during self-play training turns out to be enough to keep it from teaching itself nothing.

When two models are set up in a loop, one generating math problems and the other solving them, the generator quickly converges on a narrow set of prompts. It learns what satisfies the reward function and stops exploring. Researchers call this diversity collapse, and it quietly breaks the whole premise of autonomous curriculum learning. A team training Qwen3-4B and Qwen3-8B via a framework called R-Zero introduced vocabulary dropout: a hard, randomly-changing mask applied to the proposer model's output probabilities that prevents it from locking into fixed phrasing. Because the mask is non-stationary, the model cannot simply route around it. Diversity held up across lexical, semantic, and functional metrics throughout training.

The practical result at 8B parameters was a +4.4-point average gain on math benchmarks, with the largest improvements on competition-level problems where pattern-matching is least useful. Self-play is attractive precisely because training data costs nothing to label, but a lazy generator poisons the loop from the start. The authors frame the fix through an analogy to game rules: in chess, the rules themselves constrain the action space; here, an artificial constraint serves the same structural role.

The authors are careful to call vocabulary dropout one instantiation of a principle, not a solved problem. That framing is right: a randomly shifting mask is one collapse vector away from the next failure mode nobody has found yet.

TR

The Revision

Written by an AI system from the public sources credited above. How we write →