Skip to content
AI Atlas
PaperActive

When Does Text Inform? Benchmarking Information-Theoretic Metrics for Multimodal Time-Series Forecasting

arxiv.org/abs/2609.11282

quality89

Updated 40 min ago · first seen 12 Sept 2026

paper_01M29X34JTP7D0HNTQEPT5WQXA

Published
12 Sept 2026
T1 · 40 min ago
arXiv
2609.11282
T1 · 40 min ago
Category
cs.AI
T1 · 40 min ago

Abstract

Multimodal forecasting models that combine time series with text annotations promise richer prediction through textual context, but how do we know whether a text annotation meaningfully contributes to the forecasters prediction? This is an information-theoretic question, but to evaluate whether information-theoretic metrics can reliably measure the predictive value an annotation provides, a ground truth benchmark is needed, and none currently exist. We create a synthetic time series signal with annotations in three categories: semantically correct, incorrect, and irrelevant. Because the data generation process is fully controlled, ground-truth information content is known exactly, enabling principled evaluation of six complementary mutual information estimators (KSG, MINE, InfoNCE, CCA, PID and V-information). We show that all six estimators identify correct annotations as most informative, and are able to audit the quality of mixed text corpora, choosing the annotations that result in the best downstream forecasting results without the need for model training. Our benchmark identifies limitations of each estimator, and these are validated on seven real-world datasets, which show how estimator performance differs on weak signals. Finally, we establish practical rules for implementing these metrics for annotation auditing and fusion selection.

Authors 2

Emma Andrews, Gianmarco Mengaldo

Specification

Official page

Source:arXiv (Atom API + RSS)T1observed 40 min agohigh

Arxiv announce type
new

Source:arXiv (Atom API + RSS)T1observed 40 min agohigh

arXiv id
2609.11282

Source:arXiv (Atom API + RSS)T1observed 40 min agohigh

Categories
cs.AI, cs.IT, math.IT

Source:arXiv (Atom API + RSS)T1observed 40 min agohigh

PDF

Source:arXiv (Atom API + RSS)T1observed 40 min agohigh

Primary category
cs.AI

Source:arXiv (Atom API + RSS)T1observed 40 min agohigh

Published
12 Sept 2026

Source:arXiv (Atom API + RSS)T1observed 40 min agohigh

Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →

Provenance

Attributed facts

9

Source tiers

T19

Freshest observation

40 min ago

Conflicts

None