Tracing the Evidence Behind Zero-Shot Time-Series Forecasting: A Source-First Taxonomy and Audit Framework

A new arXiv paper (2609.21425v1) argues that zero-shot time-series forecasting should be governed as an evidence-access claim rather than a training-status condition, since a frozen language model prompted with serialized values, a time-series model pretrained on broad forecasting corpora, and a retrieval-augmented forecaster can all satisfy the no-update condition while drawing on different transferable evidence. The paper proposes a source-first taxonomy separating three primary evidence sources — frozen LLM prior reuse, parametric time-series pretraining, and retrieval-augmented external memory — from the architectures that implement them, plus four audit questions covering task interface, forecast object and scoring, prediction-time context, and resource budget. The authors' stated agenda is to make zero-shot leaderboards auditable by reporting evidence boundaries and interface assumptions alongside scores so benchmark progress reflects transferable forecasting capability rather than undisclosed changes in context, memory, or budget.

arXiv:2609.21425v1 Announce Type: new Abstract: Zero-shot time-series forecasting TSF is often described as forecasting without target-specific parameter updates, but that training-status condition does not specify what evidence the system may use. A frozen language model prompted with serialized values, a time-series model pretrained on broad forecasting corpora, and a retrieval-augmented forecaster may all satisfy the no-update condition while drawing on different transferable evidence. This paper argues that zero-shot TSF should therefore be governed as an evidence-access claim. We propose a source-first taxonomy that separates three primary evidence sources---frozen LLM prior reuse, parametric time-series pretraining, and retrieval-augmented external memory---from the architectures that implement them. After the source is identified, four additional audit questions remain: task interface, forecast object and scoring, prediction-time context, and resource budget. The resulting agenda is to make zero-shot leaderboards auditable by reporting evidence boundaries and interface assumptions alongside scores, so that benchmark progress reflects transferable forecasting capability rather than undisclosed changes in context, memory, or budget.