Skip to content
KS

RESEARCH

Publications and preprints — quantitative methods, statistical validation, and the methodology of forecasting.

papers
2
latest
2026-06-09
with doi
1

I work at the intersection of quantitative finance, statistics, and machine learning. My research interest is validation methodology — the discipline of separating a real result from an artifact of search, and importing the tools one field paid for in hard lessons into another that has not yet learned them. Every result below carries explicit epistemic labeling and says what is reproducible and what is not yet public.

Publications

2
[01] Kacper Saks · 2026 · Zenodo (v0.2.0) · erratum v0.2.1 on GitHub
Preprint

An Executable-Gate Multi-Agent Research Organization: Artifact, Case Study, and a Pre-Registered Gate-Calibration Study

Multi-agent systems are increasingly proposed as autonomous research organizations, yet their quality-control mechanisms are rarely measured. This work releases a domain-agnostic multi-agent research organization (39 agent role files plus an orchestrator) in which every role output must pass an executable gate and adversarial sign-off — nothing…

abstract & details →
cite · bibtex
bibtex
@software{saks_2026_20645678,
  author    = {Saks, Kacper},
  title     = {An Executable-Gate Multi-Agent Research Organization:
               Artifact, Case Study, and a Pre-Registered
               Gate-Calibration Study},
  month     = jun,
  year      = 2026,
  publisher = {Zenodo},
  version   = {v0.2.0},
  doi       = {10.5281/zenodo.20645678},
  url       = {https://doi.org/10.5281/zenodo.20645678}
}
[02] Kacper Saks · 2026 · Self-archived preprint
Preprint

The Validation Crisis in AI Forecasting: Deflating AGI Capability Forecasts with Quantitative-Finance Overfitting Tools

Forecasts of when artificial general intelligence will arrive increasingly shape capital allocation, regulation, and where a generation of talent is placed — yet the confidence attached to them exceeds what the methods producing them can support. We argue this gap is structural, not careless: it is the predictable consequence of fitting a model to a measured…

abstract & details →
cite · bibtex
bibtex
@misc{saks2026validation,
  title        = {The Validation Crisis in AI Forecasting: Deflating AGI Capability Forecasts with Quantitative-Finance Overfitting Tools},
  author       = {Saks, Kacper},
  year         = {2026},
  month        = jun,
  howpublished = {Self-archived preprint},
  url          = {https://kacpersaks.dev/research/validation-crisis},
  note         = {21 pp., 31 references}
}