ChatTS-14B: benchmark results
Provider: Other.
Aggregate Rating 1545 ± 45, rank #1764 of 10916 rated models, from 26 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| TS-Skill (Free-Form) - SK1+SK2+SK3 | 0.59 | Native-form score (0-1; free-form answers parsed by GPT-5.4 | 75 |
| CLIR-Bench - Cross-Variable Reasoning | 32 | Accuracy (%) on asynchronous cross-variable reasoning, 4-opt | 71.4 |
| CLIR-Bench - Time-Series Summarization | 39.67 | Accuracy (%) on clinical time-series summarization, 4-option | 71.4 |
| TS-Skill (Free-Form) - SK1+SK2 | 0.59 | Native-form score (0-1; free-form answers parsed by GPT-5.4 | 62.5 |
| TS-Skill (Free-Form) - SK1+SK3 | 0.5 | Native-form score (0-1; free-form answers parsed by GPT-5.4 | 62.5 |
| TS-Skill (MCQ) - SK1 Temporal Scale Selection | 0.72 | MCQ macro-F1 (0-1; four-option multiple choice, letter-only | 57.1 |
| CLIR-Bench | 27.14 | Macro-average accuracy (%) over the 11 tasks, 4-option multi | 50 |
| CLIR-Bench - Intervention Decision | 34.33 | Accuracy (%) on immediate intervention decision, 4-option mu | 50 |
| TS-Skill (Free-Form) - SK2+SK3 | 0.45 | Native-form score (0-1; free-form answers parsed by GPT-5.4 | 50 |
| CLIR-Bench - Intervention Response | 31.67 | Accuracy (%) on intervention response, 4-option multiple-cho | 42.9 |
| CLIR-Bench - Missingness Awareness | 26.67 | Accuracy (%) on missingness awareness (whether the observati | 42.9 |
| CLIR-Bench - Temporal Grounding | 25.33 | Accuracy (%) on temporal grounding (locate the time points o | 42.9 |
Interactive version: theaggregate.ai/model?slug=chatts-14b · How It Works · Data refreshed daily, snapshot 2026-11-08.