Researchers measure individual training example impact through paired pre-training runs
Source published ·Modelwire updated
Original coverage: arXiv cs.LG ↗·How Modelwire adds context
The development
Researchers measured rather than estimated how individual training examples shape final model behavior by running 24 paired pre-training experiments on GPT-2. By injecting a single passage at peak learning rate and comparing models trained with and without it, they isolated the causal effect of that data point on learned representations. The work bridges a methodological gap in mechanistic interpretability: most contribution estimates rely on gradient-based proxies, but this study directly quantifies what a model actually retains or forgets from a single example, offering concrete evidence for data valuation and model debugging at scale.
Modelwire’s AI-generated summary of coverage from arXiv cs.LG.
Modelwire analysis
ExplainerOur AI-generated reading of the wider context and the next developments to watch.
The key novelty is empirical rather than theoretical: by running paired pre-training experiments instead of relying on influence functions or gradient estimates, the authors can show what a model genuinely learns or discards from a single example. Most prior work infers this; this work observes it directly.
This connects to the broader mechanistic interpretability push we covered in August around understanding what models actually learn versus what we assume they learn. The Group-Calibrated distillation paper from the same day identified a teacher-verifier misalignment in how models absorb supervision. This work tackles the inverse problem: not how models learn from curated supervision, but what they retain from raw training data. Both papers share a skepticism of proxy metrics (likelihood matching in one case, gradient-based contribution estimates in the other) in favor of direct measurement of model behavior.
If follow-up work uses this single-example measurement technique to identify and remove harmful or redundant data points from large pre-training runs (OpenWebText or similar), and shows measurable downstream improvements in model behavior without retraining from scratch, that confirms the method scales to practical data curation. If the results remain confined to GPT-2 scale experiments, the practical applicability stays limited.
This interpretation is generated from the summary above and the archive coverage cited below. Our methodology · Report an error
Coverage behind this analysis
These archive entries ground the connection in our analysis. They are ordered by source publication date, with links to our coverage and the original sources.
·arXiv cs.LG
Distillation misalignment worsens with longer context in reasoning tasks
Researchers identify a fundamental misalignment in how language models learn from teacher supervision during long-context reasoning tasks. Token-level guidance from stronger models can inadvertently reward locally coherent but globally incomplete responses, while task-specific verifiers measure actual task success differently. This teacher-verifier disagreement worsens as context length increases, suggesting current distillation methods may be optimizing the…
MentionsGPT-2 · OpenWebText
How this coverage is produced
Modelwire uses AI to generate summaries and context from source headlines, snippets, and selected archive coverage. Automated checks do not verify every claim, and items are not routinely reviewed by a person before publication. Zacaria Solis operates the site. Read the linked source for the full evidence and report errors through our corrections process.
Modelwire summarizes, we don’t republish. arXiv cs.LG originally reported this story as “Learned, Then Lost: A Measured Single-Example Counterfactual in Pre-training”. The full content lives on arxiv.org. If you’re a publisher and want a different summarization policy for your work, see our takedown page.