v1.0.0 — workshop release

Code, recorded scores and project page for "Pairwise Approximation Can Select the Wrong Multi-Robot Plan" (IROS 2026 Workshop on Intelligent Information Gathering).

• analyze.py recomputes the paper's scores and plan selections from the recorded subset values in scores.json, using only the Python standard library. python -m unittest discover -s . runs the checks.
• The project page (https://www.william-teo.com/pairwise-regret/) has a replay of the env1 plans picked by the exact and pairwise scores, an interactive view of the subset values behind each score, and results for all seven maps.
• It reproduces score comparisons from recorded observations; it does not rerun the simulator. See METHODS.md for scope and limitations.

Benchmark: BYU FROST Lab indoor exploration competition, pinned at 2a2c7519.