Parallel test-time scaling samples many reasoning traces and majority-votes their answers, improving LLM accuracy but requiring traces to run to completion, incurring substantial computational overhead. We observe that probing partial traces at intermediate checkpoints can extract current answers without disrupting generation, revealing an evolving aggregate vote. Based on this observation, we introduce MARS, a margin-adversarial stopping rule that estimates which active traces are likely to change their answers and stops once the leader remains safe under a conservative bound on future vote movement. The rule separates two sources of uncertainty. It learns the trace-level switch probabilities that determine how much of the current margin is likely to be retained, while handling the harder question of where switching traces land through an adversarial bound calibrated from warmup traces. With true switch probabilities, MARS guarantees with high probability that the early-stopped answer matches the full-budget vote. In practice, a five-feature logistic model closely matches oracle switching behavior. Across three reasoning models and three competition-math benchmarks, MARS saves 25-47% of self-consistency tokens and 14-29% on top of DeepConf Online, a strong confidence-weighted baseline that already filters and truncates weak traces, while matching the accuracy of the corresponding full-budget baselines.

翻译：并行测试时扩展通过采样多条推理轨迹并对答案进行多数投票，能提升大语言模型精度，但要求所有轨迹运行至终止，产生大量计算开销。我们观察到，在中间检查点探测部分轨迹可提取当前答案而不中断生成过程，从而揭示不断演化的聚合投票结果。基于这一观察，我们提出MARS（边际对抗风险控制停止策略），该规则估计哪些活跃轨迹可能改变其答案，并在未来投票变动保守界下领先者保持安全时停止。该规则分离了两种不确定性来源：一方面学习决定当前边际保留概率的轨迹级切换概率，另一方面通过预热轨迹校准的对抗界处理切换轨迹最终归属这一更困难问题。在真实切换概率下，MARS能以高概率保证早停答案与完整预算投票结果一致。实际应用中，五特征逻辑模型能紧密逼近Oracle切换行为。在三种推理模型与三个竞赛级数学基准测试上，MARS相比完整预算基线可节省25-47%的自一致性token，并在已过滤截断弱轨迹的强置信度加权基线DeepConf Online基础上额外节省14-29%，同时保持与对应完整预算基线相当的精度。