[KongchangAI]
Benchmark

TIME

A time series forecasting evaluation benchmark used to measure a model's predictive accuracy on marginal quantiles.

Timeline (last 90 days)

Oct 22

OpenAI's first AI accelerator chip is named Jalapeño and was selected for TIME's Best Inventions of 2026 list.

Unverified50%
Oct 22

Being selected for the TIME list helps with recruiting and capital market narratives, but does not constitute technical validation.

Unverified50%
Oct 22

OpenAI Vice President of Hardware Richard Ho announced on social media the news that Jalapeño was selected for the TIME list.

Unverified50%
Oct 19

The 209M-parameter Timestral leads in the zero-shot model category on three benchmarks: GIFT-Eval, TIME and fev-bench.

Unverified50%
Oct 19

TIME is a time series forecasting evaluation system that emphasizes cross-domain and cross-sampling-frequency comprehensiveness, typically averaging over multiple datasets and multiple forecast horizons.

Unverified50%
Oct 16

Timestral-28M scores 0.678 on GIFT-Eval MASE and 0.638 on TIME MASE.

Unverified50%
Oct 16

PathFM achieves competitive marginal accuracy on the existing fev-bench and TIME benchmarks.

Unverified50%
Sep 24

Databricks Chief Information Officer Naveen Zutshi was named to TIME magazine's "Executives of the Year: Tech & Data" list.

Expired50%
Sep 24

The TIME list selects 50 leaders in the fields of technology, artificial intelligence, cybersecurity, and digital innovation.

Unverified50%
Sep 11

An authority in the AI safety field published a commentary in TIME magazine, pointing out that a cybersecurity incident involving OpenAI and Hugging Face could become a watershed moment in the AI safety domain.

Expired50%

All Facts (8)

Source Articles