InA-Probe: Instruction-Aware Active Probing for Time Series Forecasting with LLMs
arXiv:2606. 08601v1 Announce Type: new Abstract: Large Language Models (LLMs) have recently demonstrated impressive potential for time series forecasting.
The paper introduces iAmTime, a time‑series foundation model that uses instruction‑conditioned in‑context learning to adapt to tasks at inference time. iAmTime represents each episode as a structured prompt with semantic tokens that focus on specific time‑series regions, enabling the model to infer task structure from input‑output demonstrations. Trained on large real and synthetic corpora across forecasting, imputation, reconstruction, classification, anomaly detection, and source de‑mixing, iAmTime outperforms strong baselines on zero‑shot probabilistic and point forecasting while matching or exceeding performance on several non‑forecasting tasks.
arXiv:2606. 08601v1 Announce Type: new Abstract: Large Language Models (LLMs) have recently demonstrated impressive potential for time series forecasting.
TimeBraid is a family of unified models that combine pretrained language models with pretrained time‑series foundation models using interleaved global residual attention layers. The models inherit instruction following, reasoning, and continuous‑signal perception, fusing both modalities into a shared representation space for understanding and generation. The design focuses on aligning representation spaces, grounding language in temporal structure, balancing understanding with generation, and maintaining stable joint optimization, supported by 2.2 M curated series‑text pairs and 4.9 M instruction‑tuning samples. Across diverse benchmarks, TimeBraid competes with larger general‑purpose and task‑specific models.
The paper investigates when in‑context learning (ICL) can succeed even if the instruction or target task is not covered by a model’s pretraining data. It introduces a theoretical framework that distinguishes two types of distribution shift—unsupported instructions and unsupported target tasks—and shows how demonstrations can mitigate these shifts. The authors further extend the analysis to long‑horizon tasks that can be decomposed into pretraining‑supported subtasks, providing error bounds that explain how ICL can handle tasks beyond the model’s original support.
Aurora‑X is a billion‑parameter time‑series foundation model designed for extreme forecasting tasks. It employs a progressive curriculum that starts with channel‑independent pretraining, then adds cross‑variable dependencies, variable context and horizon lengths, and optional future covariates during mid‑training. A variable‑resolution post‑training stage allows adjustable temporal spans per token at inference, while a pattern‑guided mixture‑of‑experts expands capacity through sparse activation and expert specialization. An implicit quantile network head predicts arbitrary quantiles, enhancing probabilistic forecasting flexibility. Experiments on GIFT‑Eval, TIME, FEV‑Bench, TFB, and DAG‑Bench show state‑of‑the‑art performance against both pretrained TSFMs and task‑specific supervised models.
Timer‑M1 is a pretrained multivariate time‑series foundation model that learns with primitives for zero‑shot forecasting. It uses a primitive‑based data synthesis and pretraining pipeline to generate series with shared temporal and relational patterns, assembling them into multivariate samples with distinct channel roles. The model employs gated two‑dimensional Transformer blocks that dynamically allocate cross‑variate attention, achieving top performance on several large‑scale forecasting benchmarks.
arXiv:2606. 09861v1 Announce Type: cross Abstract: While Next-Token Prediction (NTP) has unified LLM pretraining, its adaptation to unbounded, continuous time series (TS) remains open.
ChronoSteer is a decoupled agentic framework that bridges large language models and time series foundation models by learning cross‑modal alignment from synthetic paired supervision. It converts textual events into revision instructions that steer a frozen time‑series model, discretizes these instructions into a compact codebook to reduce semantic divergence, and then refines the predictions with a two‑stage training strategy. The authors also release a leakage‑controlled multimodal benchmark and report a 25.8% improvement in zero‑shot prediction accuracy over the unimodal backbone.
arXiv:2610.01054v1 Announce Type: cross Abstract: In-context learning (ICL) enables language models to perform new tasks from demonstrations without weight updates. However, every ICL inference requi...
FlexCast introduces a flexible context-target learning framework for time series foundation models, reformulating forecasting as variable-length masked prediction over time series patches. By dynamically sampling visible context and masked target patches during pretraining, it exposes the model to diverse context-target configurations without sequential decoding. The same pretrained model can perform one-step forecasting, confidence-based iterative refinement, and next-token-style generation, achieving competitive or superior accuracy to strong TSFM baselines while reducing inference latency.
The paper introduces PaCTS, a method that generates instance‑adaptive latent prompts—continuous embedding tokens—to provide compact contextual information for frozen time‑series foundation models (TSFMs). These prompts are constructed from instance‑specific global statistics and refined with segment‑level temporal data, enabling the model to capture both global characteristics and local temporal variations. Experiments show that PaCTS improves forecasting performance across various context lengths and model architectures, often outperforming the same backbone with double the context while reducing inference computation, and it also offers stronger improvements and better out‑of‑distribution generalization compared to weight‑space adaptation methods.
arXiv:2609.37250v1 Announce Type: cross Abstract: World-action models (WAMs) couple future visual-state prediction with action generation. By adapting video generators or image-editing models pretrai...
The paper introduces the AGI Maze Prediction Datasets and Benchmark, a lightweight, procedurally generated grid‑world testbed for evaluating predictive models, particularly Transformers, on tasks such as per‑step transition prediction, fixed‑horizon state prediction, and sequential textual‑observation prediction. It compares byte‑level Transformer baselines with two memory‑augmented architectures, showing that a pseudo‑video spatial‑memory Transformer achieves perfect validation accuracy on selected tasks and improves sequential text‑trace prediction, while a generic auxiliary latent‑memory Transformer does not consistently help. The study highlights that structured, task‑aligned working memory can be more effective than merely increasing latent capacity, and positions the benchmark as a compact setting for testing architectures that couple textual interfaces to learned structured state.