List your company

Language Modeling on Proof-pile

2.33PerplexitySHAREDLLM
2.23922.85213.4654.0779Jun 11, 2024Oct 10, 2024Feb 9, 2025Jun 11, 2025Oct 11, 2025Feb 10, 2026Jun 12, 2026

Evaluation results

MethodLinks
SHAREDLLMContext Length=4K2026.032.33
SHAREDLLMContext Length=8K2026.032.34
CEPEContext Length=4K2026.032.38
SHAREDLLMContext Length=32K2026.032.38
CEPEContext Length=128K2026.032.39
SHAREDLLMContext Length=128K2026.032.4
CEPEContext Length=8K2026.032.43
CEPEContext Length=32K2026.032.45
LLaMA-2-32KContext Length=32K2026.032.47
YaRNContext Length=32K2026.032.47
LLaMA-3.1Context Length=128K2026.032.48
PIContext Length=32K2026.032.51
LLaMA-3.1Context Length=32K2026.032.52
LLaMA-3.1Context Length=8K2026.032.54
YaRNContext Length=8K2026.032.56
LLaMA-2-32KContext Length=8K2026.032.58
LLaMA-3.1Context Length=4K2026.032.58
LongLoRA-7B-32kModel Size=7B, Context Length=16k2025.032.6
LongLoRA-7B-32kContext Length=16k, Model Size=7B2026.062.6
PIContext Length=8K2026.032.64
OpenLLaMA-3BModel Size=3B, Context Length=2k2025.032.7
OpenLLaMA-3BContext Length=2k, Model Size=3B2026.062.7
LLaMA-2-7BModel Size=7B, Context Length=4k2025.032.72
LLaMA-2-7BContext Length=4k, Model Size=7B2026.062.72
PIContext Length=4K2026.032.77
Phi3-128kModel Size=3B, Context Length=4k2025.032.77
Phi3-128kContext Length=4k, Model Size=3B2026.062.77
YaRNContext Length=4K2026.032.82
LLaMA-2-32KContext Length=4K2026.032.87
LongLLaMA-3B†Model Size=3B, Context Length=4k2025.032.94
LongLLaMA-3BContext Length=4k, Model Size=3B, Memory=Extended Memory2026.062.94
OpenLLaMA-3BModel Size=3B, Context Length=1k2025.032.96
OpenLLaMA-3BContext Length=1k, Model Size=3B2026.062.96
YARN-128k-7bModel Size=7B, Context Length=4k2025.032.98
ERMAR (w/ 4K capacity)Model Size=3B, Context Length=2k, Memory Capacity=4K2025.032.98
YARN-128k-7BContext Length=4k, Model Size=7B2026.062.98
ERMARContext Length=2k, Model Size=3B, Memory=4K2026.062.98
MemLong (w/ 4K capacity)Model Size=3B, Context Length=16k, Memory Capacity=4K2025.032.99
MemLongContext Length=16k, Model Size=3B, Memory=4K2026.062.99
YARN-128k-7bModel Size=7B, Context Length=1k2025.033.03
ERMAR (w/ 4K capacity)Model Size=3B, Context Length=4k, Memory Capacity=4K2025.033.03
YARN-128k-7BContext Length=1k, Model Size=7B2026.063.03
ERMARContext Length=4k, Model Size=3B, Memory=4K2026.063.03
Phi3-128kModel Size=3B, Context Length=16k, GPU Configuration=Dual GPU2025.033.08
Phi3-128kContext Length=16k, Model Size=3B, GPU Setup=Dual2026.063.08
Phi3-128kModel Size=3B, Context Length=2k2025.033.11
Phi3-128kContext Length=2k, Model Size=3B2026.063.11
KGERMARContext Length=4k, Model Size=3B, Memory=4K2026.063.12
LongLLaMA-3B*Model Size=3B, Context Length=2k2025.033.15
MemLong (w/ 4K capacity)Model Size=3B, Context Length=4k, Memory Capacity=4K2025.033.15
LongLLaMA-3BContext Length=2k, Model Size=3B, Memory=No Extended Memory2026.063.15
MemLongContext Length=4k, Model Size=3B, Memory=4K2026.063.15
KGERMARContext Length=2k, Model Size=3B, Memory=4K2026.063.15
MemLong (w/ 4K capacity)Model Size=3B, Context Length=2k, Memory Capacity=4K2025.033.16
MemLongContext Length=2k, Model Size=3B, Memory=4K2026.063.16
KGERMARContext Length=16k, Model Size=3B, Memory=4K2026.063.17
MemLong-3B*Model Size=3B, Context Length=2k2025.033.18
ERMAR (w/ 4K capacity)Model Size=3B, Context Length=16k, Memory Capacity=4K2025.033.18
MemLong-3BContext Length=2k, Model Size=3B, Memory=No Extended Memory2026.063.18
ERMARContext Length=16k, Model Size=3B, Memory=4K2026.063.18
LongLLaMA-3B†Model Size=3B, Context Length=2k2025.033.22
LongLLaMA-3BContext Length=2k, Model Size=3B, Memory=Extended Memory2026.063.22
LongLoRA-7B-32kModel Size=7B, Context Length=4k2025.033.23
LongLoRA-7B-32kContext Length=4k, Model Size=7B2026.063.23
LLaMA-2-7BModel Size=7B, Context Length=1k2025.033.24
ERMAR (w/ 4K capacity)Model Size=3B, Context Length=1k, Memory Capacity=4K2025.033.24
LLaMA-2-7BContext Length=1k, Model Size=7B2026.063.24
ERMARContext Length=1k, Model Size=3B, Memory=4K2026.063.24
YARN-128k-7bModel Size=7B, Context Length=2k2025.033.29
YARN-128k-7BContext Length=2k, Model Size=7B2026.063.29
LongLoRA-7B-32kModel Size=7B, Context Length=2k2025.033.35
LongLoRA-7B-32kContext Length=2k, Model Size=7B2026.063.35
LLaMA-2-7BModel Size=7B, Context Length=2k2025.033.4
LLaMA-2-7BContext Length=2k, Model Size=7B2026.063.4
MemLong (w/ 4K capacity)Model Size=3B, Context Length=1k, Memory Capacity=4K2025.033.53
MemLongContext Length=1k, Model Size=3B, Memory=4K2026.063.53
LongLLaMA-3B*Model Size=3B, Context Length=1k2025.033.55
LongLLaMA-3B†Model Size=3B, Context Length=1k2025.033.55
LongLLaMA-3BContext Length=1k, Model Size=3B, Memory=No Extended Memory2026.063.55
LongLLaMA-3BContext Length=1k, Model Size=3B, Memory=Extended Memory2026.063.55
KGERMARContext Length=1k, Model Size=3B, Memory=4K2026.063.55
MemLong-3B*Model Size=3B, Context Length=1k2025.033.58
MemLong-3BContext Length=1k, Model Size=3B, Memory=No Extended Memory2026.063.58
LongLoRA-7B-32kModel Size=7B, Context Length=1k2025.033.68
LongLoRA-7B-32kContext Length=1k, Model Size=7B2026.063.68
Phi3-128kModel Size=3B, Context Length=1k2025.034.25
Phi3-128kContext Length=1k, Model Size=3B2026.064.25
PoSEContext Length=32K, Interpolation Method=Linear, Fine-tuned Length=4K2024.064.4
CREAMContext Length=32K, Interpolation Method=Linear, Fine-tuned Length=4K2024.064.4
PoSEContext Length=32K, Interpolation Method=YaRN, Fine-tuned Length=4K2024.064.4
CREAMContext Length=32K, Interpolation Method=YaRN, Fine-tuned Length=4K2024.064.4
CREAMContext Length=16K, Interpolation Method=Linear, Fine-tuned Length=4K2024.064.5
PoSEContext Length=16K, Interpolation Method=NTK, Fine-tuned Length=4K2024.064.5
CREAMContext Length=16K, Interpolation Method=NTK, Fine-tuned Length=4K2024.064.5
PoSEContext Length=16K, Interpolation Method=YaRN, Fine-tuned Length=4K2024.064.5
CREAMContext Length=16K, Interpolation Method=YaRN, Fine-tuned Length=4K2024.064.5
LLaMa2-7BContext Length=4K, Fine-tuned Length=4K2024.064.6
PoSEContext Length=8K, Interpolation Method=Linear, Fine-tuned Length=4K2024.064.6
PoSEContext Length=16K, Interpolation Method=Linear, Fine-tuned Length=4K2024.064.6
CREAMContext Length=8K, Interpolation Method=Linear, Fine-tuned Length=4K2024.064.6