TIMBRE: Teaching Time Series Forecasters to Read, Remember, and Reconcile

arXiv:2610.04795v1 Announce Type: new Abstract: Event-informed forecasting requires translating reports and historical responses into changes to a numerical forecast. We propose TIMBRE (Temporal Integration of Memory-Based Responses and Evidence), which combines source-aware representation, state-conditioned response transfer, and reliability-guided fusion before a frozen forecast head. A separate readout adjusts interval widths while preserving the median. In a single-seed, one-epoch development study of 13 tasks, TIMBRE improves MAE over ordinary fusion on eight tasks but over native Chronos-2 on only two. Disabling response transfer in the trained model reduces BTC and AULF MAE by 47.04% and 6.81%, respectively. These findings identify sensitivity to learned response transfer rather than a general forecasting advantage. Missing development-set scores and the absence of retrained ablations limit attribution to individual evidence mechanisms.

Related

• HARN: Hierarchical Associative Resonance Network for Event-Driven Multi-Timeframe Forecasting
• CAME: Company-Aware Evidence-Memory Experts for Interpretable Quarter-Ahead Revenue Forecasting
• DRIFT: Direct-Recursive Intervention-Conditioned Forecasting of ICU Physiological Trajectories
• BeliefGraph-JEPA: Structured Latent World Models for Action-Conditioned Time Series

Source: arXiv cs.LG | 2026-10-06