{"total":40,"items":[{"citing_arxiv_id":"2607.06504","ref_index":35,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"RMISC: A Large-scale Real-world Multivariate Corpus for Time Series Foundation Models","primary_cat":"cs.AI","submitted_at":"2026-07-07T17:04:44+00:00","verdict":"CONDITIONAL","verdict_confidence":"UNKNOWN","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A curated 142-billion-point real-world multivariate time series corpus improves zero-shot forecasting when combined with existing synthetic and univariate pretraining data across four foundation models.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2607.05291","ref_index":84,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Forecasting Realized Volatility with Time Series Foundation Models: A Comparison with Econometric Benchmarks","primary_cat":"q-fin.ST","submitted_at":"2026-07-06T16:30:41+00:00","verdict":"ACCEPT","verdict_confidence":"HIGH","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Zero-shot time series foundation models largely fail to beat econometric benchmarks for realized volatility forecasting, with only TTM achieving a narrow, calibration-driven edge.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2607.00956","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Aionoscope: Debugging Latent-State Accessibility in Time-Series Representations","primary_cat":"cs.LG","submitted_at":"2026-07-01T13:54:20+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Aionoscope shows that time-series representations recover coarse signal types reliably but expose dense latent states like phase and amplitude much less reliably, with best dense-probe R² at 0.689 versus oracle 0.999.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.30410","ref_index":81,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Beyond IID: How General Are Tabular Foundation Models, Really?","primary_cat":"cs.LG","submitted_at":"2026-06-29T14:55:52+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Tabular foundation models excel on tiny- to medium-sized IID data but are outperformed by traditional tree-based and deep learning models on non-IID, large, and high-dimensional datasets, based on evaluations across 11 models and 142 datasets in the new BeyondArena benchmark.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.24996","ref_index":3,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"From Forecasting Leaderboards to Deployment Decisions: A Fail-Closed Certification Protocol","primary_cat":"cs.LG","submitted_at":"2026-06-23T15:59:17+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Presents a fail-closed certification protocol for determining when forecasting leaderboard winners are deployment-actionable, using a traffic dataset to show friction-induced reversals and an audit to prevent overclaiming.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.13571","ref_index":44,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Existence Precedes Value: Joint Modeling of Observational Existence and Evolving States in Time Series Forecasting","primary_cat":"cs.LG","submitted_at":"2026-06-11T16:59:42+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Timeflies reformulates time series forecasting as joint inference of future observability and value estimation using coupled observation and value streams.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.13513","ref_index":2,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"CloudCons: A Comprehensive End-to-End Benchmark for Cloud Resource Consolidation","primary_cat":"cs.AI","submitted_at":"2026-06-11T16:00:51+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"CloudCons benchmark shows foundation models' superior zero-shot forecasting does not automatically yield better resource consolidation decisions, with predictive quantile choice acting as a key lever for efficiency-reliability trade-offs.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.11625","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TimeRouter: Efficient and Adaptive Routing of Time-Series Foundation Models","primary_cat":"cs.LG","submitted_at":"2026-06-10T03:39:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"TimeRouter routes among time-series foundation models via discriminative routing, selective gating and ensemble fallback, reporting SOTA LB MASE 0.6765 on GIFT-EVAL.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09420","ref_index":39,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Benchmarking Deep Time Series Models for Equity Portfolios","primary_cat":"math.OC","submitted_at":"2026-06-08T12:36:12+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Benchmark of 15 time-series architectures on equity portfolios finds no model dominates, with TransEnc-8 at 0.352 rank-1 acceptability and all promoted models showing negative net Sharpe at 20 bps costs under constraints.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09954","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Does Normalization Choice Matter for Causal Large Time-Series Models?","primary_cat":"cs.LG","submitted_at":"2026-06-08T10:51:56+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Normalization choice significantly influences training convergence and forecasting performance in causal large time-series models.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09912","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Mix, Don't Pick: Why Synthetic Corpus Composition Matters for Time Series Foundation Model Pretraining","primary_cat":"cs.LG","submitted_at":"2026-06-06T12:10:38+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Equal-weight mixture of synthetic generators matches or exceeds best single generator for time series foundation model pretraining and strengthens further with real data.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.05332","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"GITCO: Gated Inference-Time Context Optimization in TSFMs","primary_cat":"cs.AI","submitted_at":"2026-06-03T18:17:40+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"GITCO delivers +1.95% average MASE reduction on TimesFM 2.5 across 53 datasets by gated inference-time suppression of anomalous patches, capturing 89.9% of the improvement upper bound.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.01999","ref_index":48,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Why Do Time Series Models Need Long Context Windows?","primary_cat":"cs.LG","submitted_at":"2026-06-01T09:55:06+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Long input windows are required to identify the generative process in time series forecasting even for short-memory processes, and decoupling identification from forecasting improves scalability.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09861","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Time Series as Language: A Universal Tokenizer for General-Purpose Time Series Foundation Models","primary_cat":"cs.LG","submitted_at":"2026-05-31T16:04:11+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"UniTok tokenizes time series for an off-the-shelf LLM foundation model that unifies forecasting, generation, and classification through next-token prediction and training-free inference.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.01289","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Feature to Dynamics: Feature-space to Autoregression strategy for Zero-shot Time Series Forecasting","primary_cat":"cs.LG","submitted_at":"2026-05-31T15:20:14+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"FSA learns a mapping from feature space to autoregressive strategy space to improve zero-shot univariate time series forecasting over Transformer baselines under matched pretraining conditions.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.27286","ref_index":3,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Falcon-X: A Time Series Foundation Model for Heterogeneous Multivariate Modeling","primary_cat":"cs.LG","submitted_at":"2026-05-26T17:03:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Falcon-X introduces a latent prototype space with Unified Prototype Diff-Attention and Latent Entity Attention for heterogeneous multivariate time series forecasting.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.25166","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"AME-TS: Anchored Mixture-of-Experts for Time Series Forecasting","primary_cat":"cs.LG","submitted_at":"2026-05-24T16:52:48+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"AME-TS is a structure-guided sparse MoE foundation model for time series that aligns expert routing with series-level temporal descriptors to achieve strong accuracy-efficiency tradeoffs on GIFT-Eval while improving specialization stability.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.26161","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TSFMAudit: Data Contamination Auditing in Forecasting Time Series Foundation Models","primary_cat":"cs.LG","submitted_at":"2026-05-24T14:59:12+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":8.0,"formal_verification":"none","one_line_summary":"TSFMAudit detects pretraining contamination in time series foundation models via probe adaptation dynamics (faster loss drop, smaller backbone shift), tested on 6 models and 187 datasets against 10 LLM-derived baselines.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.25045","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"AION: Next-Generation Tasks and Practical Harness for Time Series","primary_cat":"cs.AI","submitted_at":"2026-05-24T12:42:12+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"AION is a time series harness using agents, skills, rules, memory, evaluation, and protocols with temporal grounding, shown in a Kaggle Store Sales case study to produce more artifacts and reviews than direct agent use.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.24381","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Assessing the Operational Viability of Foundation Models for Time Series Forecasting","primary_cat":"cs.LG","submitted_at":"2026-05-23T03:40:44+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Foundation models match or approach supervised performance in periodic and cold-start domains but lag in physically constrained systems, while a feature-based router improves accuracy and cuts inference cost versus always using one model class.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.17340","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Olivia: Harmonizing Time Series Foundation Models with Power Spectral Density","primary_cat":"cs.LG","submitted_at":"2026-05-17T09:19:22+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Olivia harmonizes time series datasets via normalized power spectral density using a Harmonizer module and resonator-based HarmonicAttention, achieving state-of-the-art zero-shot, few-shot, and full-shot forecasting on TSLib, GIFT-Eval, and GluonTS benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.12768","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"ISOMORPH: A Supply Chain Digital Twin for Simulation, Dataset Generation, and Forecasting Benchmarks","primary_cat":"stat.ML","submitted_at":"2026-05-12T21:31:32+00:00","verdict":"ACCEPT","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"ISOMORPH supplies the first public configurable digital-twin simulator for multi-echelon logistics networks together with generated datasets and zero-shot foundation-model benchmarks that exceed GIFT-Eval MASE at short horizons.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.12200","ref_index":9,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Investigating simple target-covariate relationships for Chronos-2 and TabPFN-TS","primary_cat":"cs.LG","submitted_at":"2026-05-12T14:38:25+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"TabPFN-TS captures simple target-covariate relationships more effectively than Chronos-2 in controlled experiments, especially for short horizons.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.08857","ref_index":48,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"RareCP: Regime-Aware Retrieval for Efficient Conformal Prediction","primary_cat":"cs.LG","submitted_at":"2026-05-09T10:12:07+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"RareCP improves interval efficiency for time series conformal prediction by retrieving and weighting regime-specific calibration examples while adapting to drift and maintaining coverage.","context_count":1,"top_context_role":"baseline","top_context_polarity":"baseline","context_text":"Lightning fast forecasting with statistical and econometric models. PyCon Salt Lake City, Utah, US 2022, 2022. URLhttps://github.com/Nixtla/statsforecast. [47] Roberto Neglia, Andrea Cini, Michael M. Bronstein, and Filippo Maria Bianchi. ResCP: Reser- voir Conformal Prediction for Time Series Forecasting.arXiv e-prints, art. arXiv:2510.05060, October 2025. doi: 10.48550/arXiv.2510.05060. [48] Andreas Auer, Patrick Podest, Daniel Klotz, Sebastian Böck, Günter Klambauer, and Sepp Hochreiter. Tirex: Zero-shot forecasting across long and short horizons with enhanced in-context learning, 2025. URLhttps://arxiv.org/abs/2505.23719. [49] Abdul Fatir Ansari, Lorenzo Stella, Caner Turkmen, Xiyuan Zhang, Pedro Mercado, Huibin Shen, Oleksandr Shchur, Syama Sundar Rangapuram, Sebastian Pineda Arango, Shubham"},{"citing_arxiv_id":"2605.07222","ref_index":52,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Don't Learn the Shape: Forecasting Periodic Time Series by Rank-1 Decomposition","primary_cat":"cs.LG","submitted_at":"2026-05-08T04:15:10+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"A frozen average of the last two cycles matches or exceeds eight shape-learning alternatives on 97 GIFT-Eval configurations for periodic time series forecasting.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.28149","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Explainable Load Forecasting with Covariate-Informed Time Series Foundation Models","primary_cat":"cs.LG","submitted_at":"2026-04-30T17:36:24+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Time series foundation models match the performance of specialized models for day-ahead load forecasting while providing explanations that match domain knowledge on weather and calendar effects.","context_count":1,"top_context_role":"baseline","top_context_polarity":"baseline","context_text":"claims to be able to handle covariates, recent evaluations suggest it does not effectively leverage future covariates in load forecasting applications [31]. We select Chronos-2 and TabPFN-TS, representing state-of-the- art Transformer-based and tabular foundation model paradigms, respectively, as both have shown impressive results on established forecasting benchmarks, specifically on covariate-informed tasks [1, 47]. This comparison allows us to examine whether feature impor- tance patterns are consistent across different modeling paradigms, with extension to further TSFMs left for future work. Despite their capability to generalize to unseen data, TSFMs have primarily been developed and evaluated on standard benchmark datasets. Recent work by Meyer et al."},{"citing_arxiv_id":"2604.22328","ref_index":15,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"FETS Benchmark: Foundation Models Outperform Dataset-specific Machine Learning in Energy Time Series Forecasting","primary_cat":"cs.LG","submitted_at":"2026-04-24T08:00:50+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Foundation models outperform dataset-specific machine learning in energy time series forecasting across 54 datasets in 9 categories.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"In particular, Chronos [13] in 2024 attracted attention by adapting existing language model architectures through minimal modifications - requiring only tokenization via scaling and quantization - and demonstrating strong zero-shot performance across diverse domains including finance, healthcare, nature, retail, mobility and energy [13, 14]. Since then, general-purpose benchmarks such as GIFT-EVAL [15] and FEV-Benchmark [16] have adopted similar domain coverage to systematically compare the growing number of TSFMs. By learning generalizable representations from large and diverse pretraining datasets, these models often achieve competitive performance on previously unseen data in zero-shot settings, i.e., without any task-specific training, and thus potentially providing large benefits for energy forecasting."},{"citing_arxiv_id":"2604.18521","ref_index":5,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"IDOBE: Infectious Disease Outbreak forecasting Benchmark Ecosystem","primary_cat":"cs.LG","submitted_at":"2026-04-20T17:18:18+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"IDOBE compiles over 10,000 epidemiological outbreaks into a public benchmark and shows that MLP-based models deliver the most robust short-term forecasts while statistical methods hold a slight edge pre-peak.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.00015","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TimeRFT: Stimulating Generalizable Time Series Forecasting for TSFMs via Reinforcement Finetuning","primary_cat":"eess.SP","submitted_at":"2026-04-18T06:22:26+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"TimeRFT applies reinforcement learning with multi-faceted step-wise rewards and informative sample selection to improve generalization and accuracy in TSFM adaptation beyond supervised fine-tuning.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"[9, 28, 67] requires building a dedicated model for each distinct fore- casting scenario, making them hard to attain cross-scenario gener- alization. The advent of Time Series Foundation Models (TSFMs) has overcome this bottleneck [32]. By conducting unified pretrain- ing on large-scale and heterogenous time series data, TSFMs have exhibited notable zero-shot generalization across various unseen forecasting scenarios [1]. Many domain-specific TSFMs have been developed to enhance data analytics and support decision-making, such as energy [63], healthcare [26], finance [81] and cloud [71]. Although TSFMs have shown great promise on universal zero- shot forecasting, existing TSFM research centers on unified pretrain- ing strategy, architecture design or data curation [2, 10, 42, 57, 68],"},{"citing_arxiv_id":"2604.11529","ref_index":23,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TempusBench: An Evaluation Framework for Time-Series Forecasting","primary_cat":"cs.LG","submitted_at":"2026-04-13T14:29:34+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"TempusBench is a new evaluation framework for time-series forecasting models that supplies fresh non-overlapping datasets, tasks beyond horizon and domain, consistent tuning across models, and visualization tools.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.10544","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"WaveMoE: A Wavelet-Enhanced Mixture-of-Experts Foundation Model for Time Series Forecasting","primary_cat":"cs.LG","submitted_at":"2026-04-12T09:17:19+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"WaveMoE uses a dual-path architecture with aligned time-series and wavelet tokens routed through shared experts to improve forecasting performance on diverse benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.08400","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Zero-shot Multivariate Time Series Forecasting Using Tabular Prior Fitted Networks","primary_cat":"cs.LG","submitted_at":"2026-04-09T16:00:02+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A framework recasts multivariate time series forecasting as scalar regression problems that tabular prior-fitted networks can solve zero-shot while addressing inter-channel interactions.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.06473","ref_index":1,"ref_count":2,"confidence":0.9,"is_internal_anchor":false,"paper_title":"MICA: Multivariate Infini Compressive Attention for Time Series Forecasting","primary_cat":"cs.LG","submitted_at":"2026-04-07T21:19:00+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"MICA adapts infini compressive attention to the channel dimension, enabling scalable cross-channel dependencies in Transformers and cutting forecast error by 5.4% on average versus channel-independent baselines.","context_count":1,"top_context_role":"extension","top_context_polarity":"extend","context_text":"More information on Infini-Attention is in Appendix A.4. While Infini-Attention provides a blueprint for efficient com- pression via linear attention, it was developed for long- context language modeling and does not address multivari- ate forecasting. Adapting it requires modifications. Extend- ing our prior work ( ˙Zukowska et al., 2024), our approach makes four key modifications: (1) we adapt the use of linear attention along the channel dimension to perform channel mixing and compression (2) we remove the memory update mechanism since common forecasting implementations use window-sampling during training which does not ensure sequential temporal context across forward passes required to leverage memory (Olivares et al., 2022b; Alexandrov"},{"citing_arxiv_id":"2603.22586","ref_index":1,"ref_count":2,"confidence":0.9,"is_internal_anchor":false,"paper_title":"A Foundation Model for Instruction-Conditioned In-Context Time Series Tasks","primary_cat":"cs.LG","submitted_at":"2026-03-23T21:24:41+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"iAmTime is a time-series foundation model that uses instruction-conditioned in-context learning from demonstrations to perform zero-shot adaptation on forecasting, imputation, classification, and related tasks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2602.12120","ref_index":62,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Forecasting Commencing Enrolments Under Data Sparsity: A Zero-Shot Time Series Foundation Models Framework for Higher Education Planning","primary_cat":"cs.AI","submitted_at":"2026-02-12T16:10:42+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Zero-shot TSFMs conditioned on leakage-safe covariates from Google Trends and an institutional index forecast commencing enrolments competitively with classical methods under data sparsity.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2512.20761","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TS-Arena -- A Live Forecast Pre-Registration Platform","primary_cat":"cs.LG","submitted_at":"2025-12-23T20:48:11+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"TS-Arena is a live pre-registration platform that evaluates time series forecasts on future data streams to eliminate information leakage.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2510.06063","ref_index":1,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TelecomTS: A Multi-Modal Observability Dataset for Time Series and Language Analysis","primary_cat":"cs.AI","submitted_at":"2025-10-07T15:54:34+00:00","verdict":null,"verdict_confidence":null,"novelty_score":null,"formal_verification":null,"one_line_summary":null,"context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2509.15105","ref_index":2,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Super-Linear: A Lightweight Pretrained Mixture of Linear Experts for Time Series Forecasting","primary_cat":"cs.LG","submitted_at":"2025-09-18T16:11:31+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Super-Linear introduces a pretrained MoE architecture using frequency-specialized linear experts and spectral gating for efficient general time series forecasting.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2509.06503","ref_index":24,"ref_count":2,"confidence":0.9,"is_internal_anchor":false,"paper_title":"An AI system to help scientists write expert-level empirical software","primary_cat":"cs.AI","submitted_at":"2025-09-08T10:08:36+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"ERA combines LLMs and tree search to produce expert-level empirical software that outperforms top human methods on single-cell analysis leaderboards and CDC COVID-19 forecasts.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2506.16791","ref_index":57,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TabArena: A Living Benchmark for Machine Learning on Tabular Data","primary_cat":"cs.LG","submitted_at":"2025-06-20T07:14:48+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":8.0,"formal_verification":"none","one_line_summary":"TabArena launches a dynamic, updatable benchmarking system for tabular ML that shows boosted trees remain competitive, deep learning matches them under larger budgets with ensembling, foundation models excel on small data, and cross-model ensembles advance SOTA while flagging validation overfitting.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"As we envision a living benchmarking system that will evolve over time, some limitations can be seen as future work, while others stem from fundamental trade-offs in benchmark design choices. Our (current) limitations are: (1) We use a fixed set of 200 random hyperparameter configurations to enable the study of ensemble pipelines. This prevents analyzing the variance of random hyperparameter choices [57] and studying more advanced hyperparameter optimization strategies. (2) We use a time limit per configuration; thus, our results depend on the hardware used in edge cases where the time limit is reached. Using different hardware across models (and in the future across users) reduces the comparability in such cases. (3) Our strict selection criteria for datasets makes TabArena more representative for real-world use-cases, but reduces the"}],"limit":50,"offset":0}