- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper studies sample-level effects of scaling training data during LLM fine-tuning on math reasoning. While overall test accuracy rises with more data, the authors find that 10–15% of test samples that were correctly answered with smaller training sets end up incorrect as data grows. This pattern holds across supervised fine-tuning, reinforcement learning, multiple model sizes, PEFT variants, and even test-time majority voting.
The authors show this is not driven by conflicting training samples. Instead, models trained on identical data with different random seeds learn very different functions, achieving near-identical overall accuracy but sharing little overlap in correctly answered samples. They link this to predictive multiplicity / the Rashomon effect, attributing it to a mismatch between standard next-token loss and the demands of mathematical reasoning. Ablations identify sample order and LoRA dropout as two concrete sources of run-to-run variation.
Critical strengths
- The core observation is verified across supervised fine-tuning, reinforcement learning, multiple model families and sizes, PEFT and full-parameter training, and test-time majority voting.
- Virtually all existing scaling studies only track overall accuracy, which hides the fact that adding data can degrade performance on a non-trivial subset of samples. This paper calls out this gap and demonstrates that aggregate scaling metrics can obscure important sample-level outcomes.
Critical weaknesses
- Predictive multiplicity across random seeds and the Rashomon effect are already well documented.
- The paper attributes the effect to a mismatch between next-token loss and reasoning goals, but this is presented as a plausible narrative rather than rigorously tested. There is no detailed analysis of why math reasoning specifically produces such large Rashomon sets, or what training dynamics drive previously correct samples to flip wrong as data grows.
- All main experiments use math reasoning datasets, with only one small image classification control. It is unknown whether the same pattern appears in other reasoning-heavy tasks like code generation or logical deduction, or if it is specific to math fine-tuning properties.
- The work is purely observational: it documents the problem but proposes no mitigation strategies, nor offers practical guidance for adjusting training pipelines to reduce sample-level churn.
Questions for authors
- All main experiments are on math reasoning, with only one CIFAR-10 classification control. Have you tested whether this pattern of newly incorrect samples appears in other reasoning-heavy tasks like code generation or multi-hop QA?
- The paper proposes a mismatch between next-token loss and reasoning objectives drives the effect. Are there additional analyses or ablations that more directly support this mechanism, rather than just being consistent with it?
- The work focuses entirely on documenting the phenomenon. Did you experiment with any training strategies that might reduce the rate of samples flipping from correct to incorrect as data scales? If so, how effective were they?