# Trajectory Prediction on NGSIM (Classification Metrics) Source: [Trajectory Prediction on NGSIM (Classification Metrics)]() Collection index: [llms.txt]() Data updated: 2026-03-24T12:34:38.198Z Evaluation of trajectory prediction models on the NGSIM dataset using classification-style metrics (Precision, Recall, F1, Accuracy). A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Trajectory Prediction | | Dataset (including any stated split) | NGSIM | | Metric | Precision | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Precision](): Higher is better; unit: PERCENT. - [Recall](): Higher is better; unit: PERCENT. - [F1 Score](): Higher is better; unit: PERCENT. - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Precision | Result date | Report paper | | --- | --- | --- | --- | | pythia | 99.89 | Not recorded | 68fa78881e209131cea78e6a | | OPT | 98.49 | Not recorded | 68fa78881e209131cea78e6a | | GPT-2-Large | 98.07 | Not recorded | 68fa78881e209131cea78e6a | | GPT-2 | 97.99 | Not recorded | 68fa78881e209131cea78e6a | | pythia | 97.99 | Not recorded | 68fa78881e209131cea78e6a | | OPT | 97.64 | Not recorded | 68fa78881e209131cea78e6a | | ALBERT | 97.62 | Not recorded | 68fa78881e209131cea78e6a | | BERT | 97.35 | Not recorded | 68fa78881e209131cea78e6a | | GPT-2-Medium | 97.32 | Not recorded | 68fa78881e209131cea78e6a | | XLNET | 97.19 | Not recorded | 68fa78881e209131cea78e6a | | DEBERTA | 96.94 | Not recorded | 68fa78881e209131cea78e6a | | GPT-Neo | 96.86 | Not recorded | 68fa78881e209131cea78e6a | | ROBERTA | 96.6 | Not recorded | 68fa78881e209131cea78e6a | | PALM | 96.48 | Not recorded | 68fa78881e209131cea78e6a | | bloomz | 94.01 | Not recorded | 68fa78881e209131cea78e6a | | bloomz | 93.02 | Not recorded | 68fa78881e209131cea78e6a | | flan-t5-large | 81.49 | Not recorded | 68fa78881e209131cea78e6a | | flan-t5-base | 79.96 | Not recorded | 68fa78881e209131cea78e6a | ### pythia Precision: 99.89. Result ID: 69c283e247f34ce924aa7671. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"1b"}. This record does not assert independent reproduction. ### OPT Precision: 98.49. Result ID: 69c283e247f34ce924aa7675. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"1.3b"}. This record does not assert independent reproduction. ### GPT-2-Large Precision: 98.07. Result ID: 69c283e247f34ce924aa766d. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"774M"}. This record does not assert independent reproduction. ### GPT-2 Precision: 97.99. Result ID: 69c283e247f34ce924aa766b. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"124M"}. This record does not assert independent reproduction. ### pythia Precision: 97.99. Result ID: 69c283e247f34ce924aa7670. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"410M"}. This record does not assert independent reproduction. ### OPT Precision: 97.64. Result ID: 69c283e247f34ce924aa7674. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"350M"}. This record does not assert independent reproduction. ### ALBERT Precision: 97.62. Result ID: 69c283e247f34ce924aa7665. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### BERT Precision: 97.35. Result ID: 69c283e247f34ce924aa7664. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-2-Medium Precision: 97.32. Result ID: 69c283e247f34ce924aa766c. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"355M"}. This record does not assert independent reproduction. ### XLNET Precision: 97.19. Result ID: 69c283e247f34ce924aa7669. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DEBERTA Precision: 96.94. Result ID: 69c283e247f34ce924aa7668. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-Neo Precision: 96.86. Result ID: 69c283e247f34ce924aa766a. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ROBERTA Precision: 96.6. Result ID: 69c283e247f34ce924aa7667. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### PALM Precision: 96.48. Result ID: 69c283e247f34ce924aa7666. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### bloomz Precision: 94.01. Result ID: 69c283e247f34ce924aa7673. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"1b1"}. This record does not assert independent reproduction. ### bloomz Precision: 93.02. Result ID: 69c283e247f34ce924aa7672. [Reporting paper]() Recorded evaluation configuration: {"Model Parameters":"560m"}. This record does not assert independent reproduction. ### flan-t5-large Precision: 81.49. Result ID: 69c283e247f34ce924aa766f. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### flan-t5-base Precision: 79.96. Result ID: 69c283e247f34ce924aa766e. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.