Representation Learning

Standard Overall results available

NuTime

A numerically multi-scaled time series encoder used as a frozen representation backbone.

Using NuTime for time series classification

A numerically multi-scaled time series encoder used as a frozen representation backbone.

In this benchmark, NuTime is evaluated using frozen cls representation + fitted classifier. Checkpoint information: checkpoint_bias9.pth. Check the results below for the settings and dataset scopes with published scores.

Evaluated configurations

Classifier heads are grouped under their shared pretrained model here. They remain separate leaderboard entries because they are different experimental configurations.

Organization: Microsoft Research · Peking University
Parameters: 2.4M backbone
Checkpoint used: checkpoint_bias9.pth
Input Type: Univariate + multivariate time series
Inference: Frozen CLS representation + fitted classifier

NuTime classification benchmark results

Compare accuracy and average rank only within the same evaluation setting and dataset scope. Standard uses the official training splits; 5-shot and 10-shot use different budget-feasible subsets. Rank and average rank include other models in the full comparison set. Scores are not averaged across configurations.

For comparisons with other backbones, open the foundation model classification leaderboard and use the same setting and scope. The few-shot evaluation protocol explains the label budgets.

Standard · Overall

Comparison set: 12 configurations

This Standard time series classification comparison reports 1 NuTime configuration for the overall dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 66.76%, reported by NuTime+RF with coverage 198 / 198. This configuration ranks 10 of 12 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — Standard Overall time series classification benchmark results
Configuration	Overall Rank	Overall Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	10	66.76	8.95	198 / 198

Standard · Univariate

Comparison set: 19 configurations

This Standard time series classification comparison reports 1 NuTime configuration for the univariate dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 69.39%, reported by NuTime+RF with coverage 128 / 128. This configuration ranks 16 of 19 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — Standard Univariate time series classification benchmark results
Configuration	Univariate Rank	Univariate Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	16	69.39	13.87	128 / 128

Standard · Multivariate

Comparison set: 12 configurations

This Standard time series classification comparison reports 1 NuTime configuration for the multivariate dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 61.95%, reported by NuTime+RF with coverage 70 / 70. This configuration ranks 10 of 12 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — Standard Multivariate time series classification benchmark results
Configuration	Multivariate Rank	Multivariate Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	10	61.95	9.14	70 / 70

5-shot · Overall

Comparison set: 18 configurations · Budget-feasible Low-shot subset

This 5-shot time series classification comparison reports 1 NuTime configuration for the overall dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 62.92%, reported by NuTime+RF with coverage 140 / 140. This configuration ranks 11 of 18 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — 5-shot Overall time series classification benchmark results
Configuration	Overall Rank	Overall Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	11	62.92	9.55	140 / 140

5-shot · Univariate

Comparison set: 19 configurations · Budget-feasible Low-shot subset

This 5-shot time series classification comparison reports 1 NuTime configuration for the univariate dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 64.97%, reported by NuTime+RF with coverage 112 / 112. This configuration ranks 11 of 19 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — 5-shot Univariate time series classification benchmark results
Configuration	Univariate Rank	Univariate Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	11	64.97	9.56	112 / 112

5-shot · Multivariate

Comparison set: 18 configurations · Budget-feasible Low-shot subset

This 5-shot time series classification comparison reports 1 NuTime configuration for the multivariate dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 54.73%, reported by NuTime+RF with coverage 28 / 28. This configuration ranks 17 of 18 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — 5-shot Multivariate time series classification benchmark results
Configuration	Multivariate Rank	Multivariate Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	17	54.73	12.14	28 / 28

10-shot · Overall

Comparison set: 18 configurations · Budget-feasible Low-shot subset

This 10-shot time series classification comparison reports 1 NuTime configuration for the overall dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 68.31%, reported by NuTime+RF with coverage 121 / 121. This configuration ranks 11 of 18 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — 10-shot Overall time series classification benchmark results
Configuration	Overall Rank	Overall Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	11	68.31	9.62	121 / 121

10-shot · Univariate

Comparison set: 19 configurations · Budget-feasible Low-shot subset

This 10-shot time series classification comparison reports 1 NuTime configuration for the univariate dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 70.44%, reported by NuTime+RF with coverage 97 / 97. This configuration ranks 11 of 19 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — 10-shot Univariate time series classification benchmark results
Configuration	Univariate Rank	Univariate Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	11	70.44	9.84	97 / 97

10-shot · Multivariate

Comparison set: 18 configurations · Budget-feasible Low-shot subset

This 10-shot time series classification comparison reports 1 NuTime configuration for the multivariate dataset scope. Each row shows its own coverage.

The highest reported average accuracy among this model's configurations in this setting is 59.71%, reported by NuTime+RF with coverage 24 / 24. This configuration ranks 15 of 18 in this comparison set. The result describes this configuration and setting, rather than every possible use of the pretrained model.

NuTime — 10-shot Multivariate time series classification benchmark results
Configuration	Multivariate Rank	Multivariate Accuracy (%)	Avg. Rank	Coverage
NuTime+RF	15	59.71	11.67	24 / 24

NuTime paper, code and checkpoint

Reading this model's evaluation

The method catalog describes the pretrained model; leaderboard eligibility depends on the actual coverage of each experimental configuration. Catalog inclusion does not imply complete coverage of all 198 datasets.

Compare evaluated configurations Read the evaluation and ranking rules Explore aggregate capability analysis