Time Series Foundation Models
PulseAugur coverage of Time Series Foundation Models — every cluster mentioning Time Series Foundation Models across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New framework enhances Time Series Foundation Models with fine-grained knowledge reuse
Researchers have developed a new framework called \"\method\" to improve the application of Time Series Foundation Models (TSFMs) to downstream tasks. This framework addresses challenges posed by domain shifts by decoup…
-
New study reveals systematic miscalibration in Time Series Foundation Models
A new study examines the reliability of uncertainty estimates in Time Series Foundation Models (TSFMs), which are increasingly used for probabilistic forecasting. Researchers developed a detailed evaluation framework to…
-
New QUAL framework boosts zero-shot time-series forecasting accuracy
Researchers have developed QUAL, a novel framework designed to enhance the performance of zero-shot time-series foundation models (TSFMs). QUAL operates by aligning quantiles across a pool of diverse TSFMs, using the po…
-
New framework analyzes axial attention in Time Series Foundation Models
Researchers have developed a new framework to analyze the theoretical properties of axial attention mechanisms in Time Series Foundation Models (TSFMs). This framework, an extension of the Single Location Regression met…
-
New adapter method tackles routing collapse in time series foundation models
Researchers have identified a critical issue in time series foundation models (TSFMs) where standard mixture-of-experts (MoE) adaptations fail due to "normalization-induced routing collapse." This phenomenon occurs beca…
-
New method grounds time-series models in digital twins for predictive maintenance
Researchers have developed a new method to improve the performance of time-series foundation models (TSFMs) for predictive maintenance tasks within digital twin environments. By grounding TSFMs in the topology of digita…
-
New research enhances Time-Series Foundation Models for forecasting
Four new research papers explore advancements in Time-Series Foundation Models (TSFMs), focusing on improving their forecasting capabilities. One paper introduces a method to evaluate TSFMs on sparse event data, finding…
-
New EXAONE Demand 1.0 model targets time series forecasting
Researchers have developed EXAONE Demand 1.0, a new time series foundation model specifically designed for demand forecasting. This model utilizes a unique demand-specific corpus of 11.3 million series and 48.4 billion …
-
New method uses adaptive prompts to boost time-series foundation models
Researchers have developed PaCTS, a novel method to enhance time-series foundation models (TSFMs) by using learned token embeddings as compact context surrogates. This approach generates instance-adaptive latent prompts…
-
HoliBench toolkit enables cross-platform evaluation of foundation models
A new toolkit called HoliBench has been developed to address the challenges of deploying foundation models, including large language models, on resource-constrained CPS-IoT applications. This toolkit offers a unified wo…
-
New method speeds up training for Time Series Foundation Models
Researchers have introduced Synthetic Data Distillation (SDD), a novel training objective for Time Series Foundation Models (TSFMs). SDD enhances pre-training by comparing TSFM outputs to the conditional forecast distri…
-
AI in Finance: Progress Limited for Consistent Profitability
A new research paper reviews the current state of artificial intelligence in equity and crypto markets, examining its progress from data analysis to automated investing. While AI has shown advancements in prediction, te…
-
Research reveals universal redundancies in Time Series Foundation Models
A new research paper published on arXiv details findings about universal redundancies in Time Series Foundation Models (TSFMs). The study, conducted by Anthony Bao and others, utilized mechanistic interpretability tools…
-
Decathlon adopts Time Series Foundation Models for scalable demand forecasting
Decathlon is transitioning its demand forecasting operations to Time Series Foundation Models (TSFMs) to address the scalability and efficiency challenges of managing forecasts for tens of thousands of products across m…
-
New 'Living Benchmark' LiveHouse-TS Challenges Static Time Series Model Evaluation
Researchers have introduced LiveHouse-TS, a novel benchmark infrastructure designed to evaluate Time Series Foundation Models (TSFMs) in dynamic, real-world conditions. Unlike traditional static benchmarks, LiveHouse-TS…
-
New OATS strategy enhances Time Series Foundation Models with dynamic data augmentation
Researchers have developed OATS, a novel online data augmentation strategy for Time Series Foundation Models (TSFMs). This method dynamically generates synthetic data tailored to specific training stages, using valuable…
-
New methods accelerate LLM inference with speculative decoding · 7 sources tracked
Researchers are developing new methods to accelerate the inference speed of large language models (LLMs) through speculative decoding. DARTree and SPADE are two such approaches, with DARTree focusing on tree-based specu…
-
New regularization technique combats suboptimal collapse in time series models
Researchers have introduced a new technique called Ground-Truth Neighborhood Regularization (GTN-R) to improve the performance of time series foundation models (TSFMs) when using reinforcement learning (RL) for post-tra…
-
New framework personalizes federated adaptation for time-series models
Researchers have developed a personalized federated sparse adaptation framework for time-series foundation models (TSFMs), aiming to improve energy forecasting by addressing the non-IID nature of private, distributed me…
-
LLM as Forecasting Planner framework integrates LLMs with TSFMs for improved forecasting
Researchers have developed a novel framework called LLM as Forecasting Planner (rc) that integrates large language models (LLMs) with time-series foundation models (TSFMs) for improved text-conditioned forecasting. This…