Language Modeling
Benchmarks
DatasetSOTA methodMetricTrendResultsLast updated
13.13Perplexity
12
Feb 26, 2026
2.74Final Loss
12
Feb 26, 2026
3.68Performance (8K Context)
12
Feb 26, 2026
5.44Perplexity (Baseline)
12
May 14, 2026
10.9Perplexity (PPL)
12
Feb 26, 2026
17.12Perplexity
12
Feb 26, 2026
15.59PPL
12
Feb 26, 2026
40.2Perplexity
12
Feb 26, 2026
94Perplexity
12
Feb 26, 2026
17.69PPL
11
Jun 26, 2026
15.52PPLmed
11
Jun 18, 2026
6.63Perplexity
11
Jun 15, 2026
2.64Perplexity (PPL)
11
May 29, 2026
14.05Perplexity (PPL)
11
May 14, 2026
4.05Inference Speedup
11
Jun 16, 2026
11.22Perplexity (PPL)
11
Apr 16, 2026
12.4Training Time (s)
11
Jun 23, 2026
8.937Perplexity
11
Mar 4, 2026
18.398Perplexity
11
Mar 4, 2026
14.407Perplexity
11
Mar 4, 2026
2.61Perplexity
11
Feb 26, 2026
20.96NELBO PPL
11
Feb 26, 2026
1.863Loss
11
Mar 12, 2026
0.976Bits Per Token (BPT)
11
Feb 26, 2026
56.5Perplexity
11
Feb 26, 2026