Source record

Published by arXiv cs.MA (T1 source). The original is at https://arxiv.org/abs/2610.13431.

Pipeline notes

The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.

Analysis of 928 forecasts for 232 assets across four production batches of iPulse AI finds that advisor agreement does not generally indicate forecast reliability. Mean absolute error rises with disagreement at the same rate as a no-change baseline; low-volatility filtering outperforms low-disagreement filtering.

Separates three conflated questions (asset risk, directional agreement, magnitude precision) using empirical production data. Useful for anyone building multi-agent pipelines that treat consensus as a confidence proxy.

Cited by

No citations on record.