# Reverse Supervision at Scale: Exponential Search Meets the Economics of Annotation

arXiv: https://arxiv.org/abs/2510.10446 · HTML: https://mlsanity.com/paper/2510.10446/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 24 references that are in the map.
- First posted on arXiv 12 Oct 2025.

## Builds on (most cited first, 6 of 6)

- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [Parameter-Efficient Transfer Learning for NLP](https://mlsanity.com/paper/1902.00751/) (2019)
- [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022)
- [Learning Transferable Features with Deep Adaptation Networks](https://mlsanity.com/paper/1502.02791/) (2015)
- [Model Cards for Model Reporting](https://mlsanity.com/paper/1810.03993/) (2018)
- [Data Programming: Creating Large Training Sets, Quickly](https://mlsanity.com/paper/1605.07723/) (2016)

## Abstract (arXiv)

We analyze a reversed-supervision strategy that searches over labelings of a large unlabeled set (B) to minimize error on a small labeled set (A). The search space is (2n), and the resulting complexity remains exponential even under large constant-factor speedups (e.g., quantum or massively parallel hardware). Consequently, arbitrarily fast -- but not exponentially faster -- computation does not obviate the need for informative labels or priors. In practice, the machine learning pipeline still requires an initial human contribution: specifying the objective, defining classes, and providing a seed set of representative annotations that inject inductive bias and align models with task semantics. Synthetic labels from generative AI can partially substitute provided their quality is human-grade and anchored by a human-specified objective, seed supervision, and validation. In this view, generative models function as label amplifiers, leveraging small human-curated cores via active, semi-supervised, and self-training loops, while humans retain oversight for calibration, drift detection, and failure auditing. Thus, extreme computational speed reduces wall-clock time but not the fundamental supervision needs of learning; initial human (or human-grade) input remains necessary to ground the system in the intended task.