Language Modeling
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
14.8Perplexity (PPL)
28
May 28, 2026
9.44Perplexity
28
Mar 4, 2026
4.11Perplexity (PPL)
27
Jun 4, 2026
5.92NPM
27
Mar 5, 2026
15.8Perplexity (PPL)
27
Feb 26, 2026
2.597Loss
26
May 25, 2026
3.078Validation Loss
26
May 20, 2026
8.005Perplexity
26
May 19, 2026
1.71Perplexity (PPL)
26
Jun 4, 2026
5.84PPL
26
Feb 26, 2026
8.6Perplexity (PPL)
26
Feb 26, 2026
7.2Perplexity (PPL)
26
Feb 26, 2026
14.71Perplexity
26
Feb 26, 2026
9.84Validation PPL
26
Feb 26, 2026
93.4Accuracy Retention
25
Jun 2, 2026
5.45Perplexity (PPL)
25
May 22, 2026
43.3Delta (%)
25
Apr 7, 2026
0Delta PPL
25
Apr 7, 2026
28.53Perplexity
25
May 14, 2026
0.54PPL (CommonCrawl)
25
Jun 4, 2026
2.524Train Loss
25
Jul 3, 2026
9.8Perplexity (PPL)
25
Feb 26, 2026
55.9Accuracy (Mean)
25
Feb 26, 2026
7.533Perplexity
25
Jun 24, 2026
9.58Perplexity
24
Jun 4, 2026