From the 1 of 2 linked papers with an AI index.
2 papers
cs.LG2026
Temperature Scaling Is Not Enough: Calibration Gaps Under Human Label Distributions
Wisdom Dogah
The paper investigates how temperature scaling, a common post‑hoc calibration method, performs when training labels are soft or distributional rather than one‑hot, finding consiste…
cs.AI2026
Traxia: A Framework for Verifiable, Agent-Native Scientific Publishing
Wisdom Dogah
Verifiability, attribution, and reproducibility are foundational requirements of scientific knowledge, yet current publishing infrastructure does not enforce them at scale. We intr…