Guided Self-Evolving LLMs with Minimal Human Supervision
This addresses the challenge of enabling stable and controllable AI self-evolution with reduced human oversight, which is incremental as it builds on existing self-play methods by adding lightweight guidance.
The paper tackles the problem of unguided self-evolving LLMs plateauing or degrading due to issues like concept drift and bias reinforcement, by introducing R-Few, a guided Self-Play Challenger-Solver framework with minimal human supervision. The result shows consistent iterative improvements, such as Qwen3-8B-Base improving by +3.0 points over R-Zero on math tasks and matching performance with a model trained on 20 times more human data.
AI self-evolution has long been envisioned as a path toward superintelligence, where models autonomously acquire, refine, and internalize knowledge from their own learning experiences. Yet in practice, unguided self-evolving systems often plateau quickly or even degrade as training progresses. These failures arise from issues such as concept drift, diversity collapse, and mis-evolution, as models reinforce their own biases and converge toward low-entropy behaviors. To enable models to self-evolve in a stable and controllable manner while minimizing reliance on human supervision, we introduce R-Few, a guided Self-Play Challenger-Solver framework that incorporates lightweight human oversight through in-context grounding and mixed training. At each iteration, the Challenger samples a small set of human-labeled examples to guide synthetic question generation, while the Solver jointly trains on human and synthetic examples under an online, difficulty-based curriculum. Across math and general reasoning benchmarks, R-Few achieves consistent and iterative improvements. For example, Qwen3-8B-Base improves by +3.0 points over R-Zero on math tasks and achieves performance on par with General-Reasoner, despite the latter being trained on 20 times more human data. Ablation studies confirm the complementary contributions of grounded challenger training and curriculum-based solver training, and further analysis shows that R-Few mitigates drift, yielding more stable and controllable co-evolutionary dynamics.