English
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Bilingual Response Generation | English en | QA-F154.1 | 24 | |
| Logical Reasoning | English | BPA98.8 | 24 | |
| Text-To-Speech | English (test) | WER0.0165 | 21 | |
| Helpfulness prediction | English user-disjoint | Main Score63.6 | 20 | |
| General Language Evaluation | English lm-evaluation-harness | ARC Easy Acc (Norm)0.819 | 16 | |
| Dependency Parsing | English (en) (test) | LAS95.33 | 16 | |
| Incremental BPE Tokenization | English | Median End-to-end CPU Time (s)0.925 | 15 | |
| Unsupervised Constituency Parsing | English SPMRL (test) | S-F169.7 | 15 | |
| Automatic Speech Recognition | English | WER13.35 | 12 | |
| BPE Tokenization | English | Speedup Factor3.13 | 12 | |
| Implicit Discourse Relation classification | English (test) | Precision62 | 12 | |
| LLM-as-a-judge Evaluation | English (test) | Overall Score72.7 | 11 | |
| Tokenization | English eng | NSL Score1.27 | 10 | |
| Morphological Alignment | English 300 MB Corpora | Morph. Score64.4 | 9 | |
| Bias-Penalized Accuracy Evaluation | English | Bias-Penalized Accuracy (BPA)98.78 | 9 | |
| Tokenization Efficiency | English | Bytes per Token3.65 | 6 | |
| Speech-to-Singing conversion | English (test) | LSD2.512 | 6 | |
| RST Parsing | English | Span Score88.2 | 6 | |
| ASR error correction | English (en) | WER32 | 5 | |
| Text-to-Speech | English high-resource | MOS4.61 | 5 | |
| audio-driven facial animation | English (test) | MSE0.01 | 5 | |
| Dysarthria Detection | English | Accuracy96.57 | 5 | |
| Alzheimer's Detection | English subset | Accuracy96.73 | 5 | |
| Depression Detection | English | Accuracy97.04 | 5 | |
| Speech Intelligibility Assessment | English | Absolute Kendall's Tau0.768 | 5 |
Showing 25 of 44 rows