\begin{table}[t]
\caption{Foundation-model, conformal, and emerging 2026 evidence.}\label{tab:recent2026_fm}
\small
\begin{tabular}{@{}L{0.22\textwidth}L{0.16\textwidth}L{0.24\textwidth}L{0.28\textwidth}@{}}
\toprule
Paper & Evidence status & Main setting & Role in this review \\
\midrule
LANE \citep{sosea2026lane} & Published, ICLR 2026 & Fine-grained text classification & Uses label relationships and label-aware margins to detect semantic label noise in text \\
RRM \citep{chen2026rockafellian} & Published, ICLR 2026 & Loss reweighting & Provides an architecture-independent wrapper for loss reweighting under noisy labels \\
Noise-aware generalization \citep{wang2026nag} & Published, ICLR 2026 & Domain generalization plus noise & Shows that LNL methods may confuse domain shift with label noise, motivating joint evaluation \\
RE-PO \citep{cao2026repo} and Semi-DPO \citep{liu2026semidpo} & Published, ICLR 2026 & Preference/noisy alignment labels & Extends noisy-supervision reasoning to LLM and diffusion preference optimization \\
CMRM \citep{shi2026cmrm} & Published, AISTATS 2026 & Conformal robust learning & Adds a quantile-calibrated margin envelope without privileged clean data or transition matrices \\
NCSAM \citep{xu2026ncsam} and OccNL \citep{li2026occnl} & NCSAM: arXiv preprint; OccNL: IROS 2026 acceptance reported by the authors & Optimization and 3D perception & Indicate emerging work on flatness-aware noisy optimization and task-specific 3D occupancy noise \\
\botrule
\end{tabular}
\end{table}