one figure per paper
[01]
ICLR 2026 Workshop on Reasoning and Planning for LLMs
Shows that after solver-guided repair, the dominant failure mode is not contradiction but answers that violate a still-satisfiable maintained state.
816 problems · 4 open-weight models · 68–95% of residual errors are satisfiable drift
[02]
MathAI 2026
Derives a finite-sample transfer bound for nonlinear Fourier Neural Operators that separates complexity, discretization, distribution shift, and mesh aliasing.
Four channels · Darcy, Helmholtz, and Burgers diagnostics · aliasing made explicit
[03]
IEEE ICBCB 2026
Uses minimal neuron-level features to recover species-level signatures while showing how missing-data handling can create reconstruction artifacts.
176,914 neurons · 5 species · 92.2% Random Forest accuracy
[04]
SemEval 2026 Task 6
A confidence-gated verification pipeline for political evasion detection where opposing verification passes both route corrections through the Ambivalent class.
85% Macro F1 · rank 2 · +5.2 to +21.7 over zero-shot
[05]
SemEval 2026 Task 4
Treats vote distribution as a difficulty signal, routing easy cases through supermajority voting and reserving symbolic checks for the ambiguous tail.
72.75% accuracy · 11/47 teams · 74% resolved without escalation