German NLG 🇩🇪

gpt-4-0613 (few-shot, val)

unknown

100

8220

True

597 ± 197 / 93 ± 33

1.28

69.48 ± 2.32 / 54.66 ± 2.17

64.91 ± 1.86 / 75.96 ± 1.59

50.23 ± 4.16 / 74.54 ± 2.10

33.17 ± 1.86 / 65.14 ± 1.53

66.35 ± 0.59 / 19.81 ± 1.46

71.92 ± 2.22 / 78.79 ± 1.65

82.85 ± 1.48 / 86.99 ± 1.12

12.10.0

meta-llama/Llama-3.1-405B-Instruct-FP8 (few-shot)

405869

128

131072

True

799 ± 246 / 112 ± 38

1.34

72.87 ± 1.32 / 69.43 ± 1.71

60.79 ± 2.05 / 71.73 ± 2.13

50.25 ± 1.78 / 74.33 ± 0.79

28.34 ± 1.37 / 57.73 ± 1.57

66.17 ± 0.18 / 18.11 ± 0.49

74.98 ± 0.83 / 81.27 ± 0.62

83.48 ± 1.05 / 87.58 ± 0.78

14.0.4

meta-llama/Meta-Llama-3-70B (few-shot, val)

70554

128

8221

True

312 ± 55 / 177 ± 51

1.34

69.04 ± 2.51 / 61.10 ± 3.39

63.51 ± 2.57 / 75.01 ± 1.74

37.41 ± 2.43 / 67.63 ± 1.05

38.29 ± 3.54 / 69.69 ± 2.78

70.00 ± 1.41 / 29.78 ± 3.58

64.93 ± 2.76 / 73.71 ± 2.07

66.54 ± 4.29 / 74.45 ± 3.24

12.7.0

Qwen/QwQ-32B-Preview (few-shot)

32764

152

32768

True

2,258 ± 1,221 / 198 ± 67

1.35

70.66 ± 1.22 / 61.82 ± 2.45

59.33 ± 2.00 / 69.45 ± 2.43

54.53 ± 2.50 / 76.97 ± 1.30

32.66 ± 1.90 / 62.45 ± 2.73

65.00 ± 0.76 / 16.07 ± 1.83

67.84 ± 0.70 / 75.71 ± 0.55

79.00 ± 1.09 / 83.96 ± 0.86

14.0.4

gpt-4-1106-preview (few-shot, val)

unknown

100

128000

True

576 ± 221 / 81 ± 28

1.35

68.94 ± 2.50 / 44.89 ± 2.85

60.47 ± 2.94 / 72.79 ± 1.90

51.26 ± 4.94 / 74.76 ± 2.45

30.04 ± 1.30 / 58.77 ± 1.50

63.62 ± 0.65 / 13.08 ± 0.87

73.80 ± 2.03 / 80.31 ± 1.54

83.93 ± 2.38 / 87.85 ± 1.80

12.6.1

12.10.2

12.6.1

12.10.2

Qwen/Qwen2.5-72B-Instruct (few-shot)

72706

152

32768

True

1,219 ± 412 / 158 ± 53

1.44

68.94 ± 0.70 / 51.96 ± 2.24

58.78 ± 1.77 / 72.15 ± 1.29

52.66 ± 1.92 / 75.17 ± 0.93

27.62 ± 0.83 / 57.71 ± 0.89

63.32 ± 0.06 / 12.00 ± 0.10

70.49 ± 0.48 / 77.80 ± 0.36

81.41 ± 0.90 / 85.98 ± 0.68

14.0.4

gpt-4o-2024-05-13 (few-shot, val)

unknown

200

127903

True

916 ± 329 / 114 ± 38

1.44

69.99 ± 1.63 / 45.58 ± 2.42

54.82 ± 2.19 / 68.42 ± 1.67

43.66 ± 5.67 / 64.64 ± 4.65

30.06 ± 1.04 / 60.77 ± 1.11

63.80 ± 0.60 / 13.87 ± 1.05

74.13 ± 1.60 / 80.59 ± 1.22

88.18 ± 1.79 / 91.13 ± 1.34

12.10.0

12.10.2

12.10.0

12.10.2

meta-llama/Llama-3.1-70B-Instruct (few-shot)

70554

128

131072

True

1,409 ± 457 / 186 ± 63

1.44

72.11 ± 1.42 / 54.20 ± 2.19

59.87 ± 1.63 / 72.22 ± 1.34

46.82 ± 1.39 / 72.14 ± 0.64

30.64 ± 1.27 / 57.19 ± 1.78

66.20 ± 0.16 / 18.95 ± 0.43

68.80 ± 0.65 / 76.52 ± 0.52

73.49 ± 1.15 / 79.81 ± 0.87

14.0.3

meta-llama/Llama-3.3-70B-Instruct (few-shot)

70554

128

131072

True

1,353 ± 443 / 180 ± 61

1.55

70.47 ± 1.16 / 51.68 ± 2.18

60.55 ± 2.01 / 72.05 ± 1.69

50.13 ± 1.53 / 72.87 ± 0.69

20.42 ± 0.81 / 47.44 ± 1.06

65.20 ± 0.10 / 15.59 ± 0.30

69.32 ± 0.80 / 76.95 ± 0.62

78.33 ± 0.90 / 83.58 ± 0.69

14.0.3

upstage/SOLAR-10.7B-v1.0 (few-shot)

10732

32

4096

True

3,780 ± 906 / 799 ± 261

1.56

68.11 ± 1.32 / 56.25 ± 1.65

59.79 ± 1.60 / 71.47 ± 1.54

35.45 ± 3.06 / 66.13 ± 1.28

37.27 ± 1.23 / 68.54 ± 1.94

69.31 ± 0.95 / 28.82 ± 2.24

41.20 ± 0.85 / 55.90 ± 0.66

72.65 ± 0.69 / 79.18 ± 0.58

12.5.3

google/gemma-2-27b-it (few-shot)

27227

256

8224

True

1,516 ± 257 / 480 ± 148

1.57

64.13 ± 1.65 / 55.46 ± 2.00

60.28 ± 1.75 / 73.37 ± 1.43

46.69 ± 1.99 / 71.96 ± 0.73

28.54 ± 1.38 / 59.38 ± 1.85

67.87 ± 0.99 / 24.26 ± 2.65

59.43 ± 0.82 / 69.41 ± 0.61

71.59 ± 1.49 / 78.30 ± 1.23

13.0.0

gpt-4-1106-preview (zero-shot, val)

unknown

100

8191

True

436 ± 152 / 57 ± 21

1.60

57.47 ± 2.20 / 36.29 ± 1.91

58.67 ± 2.30 / 71.20 ± 1.34

54.55 ± 3.18 / 75.08 ± 1.71

27.02 ± 1.33 / 53.97 ± 1.74

62.91 ± 0.22 / 11.70 ± 0.28

69.23 ± 2.31 / 77.07 ± 1.72

71.17 ± 2.36 / 77.66 ± 1.84

14.0.3

meta-llama/Meta-Llama-3-70B-Instruct (few-shot, val)

70554

128

8221

True

1,673 ± 583 / 275 ± 85

1.61

75.20 ± 2.15 / 64.06 ± 3.60

54.38 ± 3.31 / 68.02 ± 2.21

36.59 ± 4.24 / 67.36 ± 1.77

26.90 ± 2.67 / 58.28 ± 2.22

68.63 ± 1.20 / 26.12 ± 3.00

62.69 ± 1.83 / 71.95 ± 1.40

69.18 ± 2.73 / 76.41 ± 2.08

12.7.0

gpt-4o-2024-05-13 (zero-shot, val)

unknown

200

8191

True

637 ± 306 / 92 ± 31

1.65

67.18 ± 1.47 / 35.10 ± 1.57

50.12 ± 2.00 / 65.79 ± 1.25

44.98 ± 4.97 / 71.63 ± 2.39

27.01 ± 1.67 / 55.19 ± 1.49

63.10 ± 0.20 / 11.09 ± 0.30

74.22 ± 1.79 / 80.74 ± 1.32

69.31 ± 1.88 / 76.76 ± 1.46

14.0.3

ThatsGroes/gemma-2-27b-it-FP8-Dynamic (few-shot)

28411

256

4096

True

3,633 ± 1,236 / 777 ± 220

1.66

73.31 ± 1.01 / 59.78 ± 2.14

58.02 ± 1.62 / 71.81 ± 1.10

45.12 ± 0.99 / 70.92 ± 0.38

24.67 ± 1.03 / 55.19 ± 1.27

64.40 ± 0.06 / 15.05 ± 0.16

58.70 ± 0.65 / 68.95 ± 0.50

68.90 ± 1.33 / 76.26 ± 1.04

14.0.4

google/gemma-2-9b (few-shot)

9242

256

8224

True

2,038 ± 406 / 566 ± 172

1.67

47.39 ± 1.96 / 34.14 ± 1.70

62.89 ± 2.21 / 74.51 ± 1.63

37.22 ± 4.67 / 67.37 ± 2.22

39.11 ± 2.14 / 67.67 ± 2.90

70.44 ± 1.18 / 30.83 ± 2.69

53.28 ± 1.15 / 64.86 ± 0.90

54.51 ± 2.74 / 64.51 ± 2.33

13.0.0

google/gemma-2-9b-it (few-shot)

9242

256

8224

True

2,062 ± 397 / 589 ± 178

1.70

50.35 ± 2.81 / 39.72 ± 2.62

58.60 ± 2.21 / 71.12 ± 1.12

45.78 ± 1.55 / 70.64 ± 0.61

30.46 ± 0.98 / 60.56 ± 2.25

67.99 ± 1.12 / 24.07 ± 2.91

53.61 ± 0.60 / 64.92 ± 0.47

68.16 ± 0.75 / 75.71 ± 0.65

13.0.0

152334H/miqu-1-70b-sf (few-shot, val)

68977

32

32793

True

2,126 ± 676 / 319 ± 104

1.71

65.19 ± 2.58 / 56.17 ± 3.57

59.80 ± 2.15 / 71.98 ± 1.46

41.86 ± 5.44 / 69.70 ± 2.31

25.51 ± 3.79 / 63.19 ± 2.48

66.80 ± 0.76 / 20.65 ± 1.59

55.18 ± 3.35 / 66.48 ± 2.49

61.68 ± 3.07 / 70.78 ± 2.37

12.7.0

meta-llama/Llama-2-70b-hf (few-shot, val)

68977

32

4125

True

1,892 ± 650 / 318 ± 105

1.71

63.71 ± 2.43 / 57.08 ± 2.70

58.17 ± 2.51 / 71.34 ± 1.62

36.33 ± 5.00 / 64.51 ± 3.38

36.06 ± 2.89 / 69.62 ± 2.81

69.82 ± 1.03 / 30.79 ± 2.46

46.44 ± 2.49 / 59.88 ± 1.74

48.89 ± 2.85 / 61.33 ± 2.06

12.7.0

mistralai/Mistral-Nemo-Instruct-2407 (few-shot)

12248

131

1024032

True

7,095 ± 2,193 / 1,063 ± 344

1.74

66.27 ± 1.13 / 49.86 ± 1.62

57.70 ± 2.68 / 71.63 ± 1.94

35.54 ± 4.25 / 60.12 ± 5.53

34.45 ± 1.61 / 66.90 ± 2.58

68.65 ± 1.15 / 25.16 ± 2.66

47.05 ± 1.12 / 60.01 ± 0.86

52.68 ± 3.39 / 63.07 ± 2.88

13.0.0

gpt-4o-mini-2024-07-18 (few-shot, val)

unknown

200

8191

True

784 ± 310 / 95 ± 28

1.77

65.51 ± 1.91 / 48.61 ± 3.73

55.16 ± 3.05 / 69.81 ± 1.96

44.60 ± 3.19 / 67.35 ± 2.53

21.87 ± 3.80 / 59.67 ± 2.67

63.50 ± 0.20 / 11.55 ± 0.36

60.83 ± 2.50 / 70.31 ± 1.91

67.98 ± 3.25 / 75.39 ± 2.59

14.0.0

14.0.1

skole-gpt-mixtral (few-shot)

unknown

32

32768

False

3,583 ± 977 / 686 ± 231

1.78

57.82 ± 1.85 / 43.22 ± 2.36

59.45 ± 1.95 / 72.95 ± 1.33

36.75 ± 2.58 / 67.88 ± 1.50

33.55 ± 2.29 / 63.95 ± 2.83

69.37 ± 0.84 / 28.13 ± 2.16

50.96 ± 0.60 / 63.17 ± 0.44

39.28 ± 2.50 / 52.23 ± 2.22

13.0.0

gpt-4o-mini-2024-07-18 (zero-shot, val)

unknown

200

8191

True

908 ± 303 / 96 ± 36

1.82

59.19 ± 1.52 / 30.16 ± 1.22

52.66 ± 3.86 / 68.10 ± 2.51

46.66 ± 4.45 / 72.82 ± 2.13

26.02 ± 1.78 / 51.63 ± 1.54

63.30 ± 0.18 / 11.78 ± 0.24

58.21 ± 2.60 / 68.09 ± 2.09

65.19 ± 2.42 / 73.28 ± 1.95

14.0.1

14.0.3

CohereForAI/c4ai-command-r-08-2024 (few-shot)

32296

256

131072

False

1,909 ± 646 / 248 ± 84

1.85

68.62 ± 1.15 / 49.68 ± 3.91

56.20 ± 2.11 / 69.68 ± 1.83

43.36 ± 1.37 / 67.19 ± 2.03

28.50 ± 1.00 / 57.50 ± 1.84

64.52 ± 0.06 / 13.88 ± 0.25

50.16 ± 1.00 / 62.25 ± 0.74

51.13 ± 2.01 / 62.67 ± 1.67

14.0.4

gpt-3.5-turbo-0613 (few-shot, val)

unknown

100

4095

True

921 ± 293 / 113 ± 37

1.95

61.50 ± 2.96 / 46.22 ± 3.41

55.50 ± 2.58 / 68.96 ± 2.00

38.96 ± 4.39 / 68.89 ± 2.54

30.20 ± 1.59 / 56.58 ± 1.78

64.90 ± 0.22 / 15.99 ± 0.32

35.39 ± 3.89 / 51.41 ± 2.98

56.88 ± 2.50 / 66.76 ± 2.02

0.0.0

mistralai/Mixtral-8x7B-Instruct-v0.1 (few-shot)

46703

32

32768

True

5,535 ± 1,837 / 760 ± 256

2.03

58.07 ± 0.94 / 40.30 ± 1.75

52.18 ± 1.33 / 65.98 ± 1.66

41.45 ± 1.66 / 68.13 ± 0.98

26.87 ± 1.60 / 59.41 ± 2.41

64.10 ± 0.11 / 13.39 ± 0.19

47.93 ± 0.83 / 60.80 ± 0.60

51.72 ± 1.42 / 63.09 ± 1.17

14.0.4

nvidia/Llama-3.1-Nemotron-70B-Instruct-HF (few-shot)

70554

128

131072

True

1,208 ± 412 / 156 ± 53

2.04

61.98 ± 0.95 / 33.82 ± 1.10

57.07 ± 1.62 / 70.61 ± 1.27

45.61 ± 1.62 / 71.65 ± 0.69

24.89 ± 0.88 / 54.36 ± 1.28

43.28 ± 0.70 / 1.38 ± 0.31

66.68 ± 0.93 / 74.94 ± 0.72

75.48 ± 1.28 / 81.42 ± 1.00

14.0.4

mistralai/Ministral-8B-Instruct-2410 (few-shot)

8020

131

32768

True

1,302 ± 323 / 253 ± 86

2.05

60.50 ± 1.22 / 40.72 ± 1.61

50.39 ± 2.45 / 66.39 ± 1.62

30.86 ± 1.37 / 53.78 ± 1.61

30.53 ± 1.21 / 58.26 ± 1.93

64.03 ± 0.05 / 12.90 ± 0.19

39.61 ± 0.90 / 54.41 ± 0.67

60.04 ± 1.94 / 69.43 ± 1.54

14.1.2

VAGOsolutions/SauerkrautLM-7b-LaserChat (few-shot)

7242

32

4096

False

4,413 ± 1,265 / 551 ± 184

2.07

62.37 ± 0.98 / 39.99 ± 2.90

49.59 ± 1.99 / 65.96 ± 1.43

35.57 ± 1.53 / 66.87 ± 0.69

31.74 ± 0.72 / 63.96 ± 1.35

64.55 ± 0.08 / 14.48 ± 0.22

34.68 ± 0.90 / 50.71 ± 0.72

49.51 ± 1.27 / 61.62 ± 1.04

14.1.2

meta-llama/Meta-Llama-3-8B-Instruct (few-shot)

8030

128

8192

True

1,483 ± 377 / 287 ± 97

2.07

68.18 ± 0.95 / 57.72 ± 1.15

58.33 ± 2.83 / 69.31 ± 3.16

29.12 ± 3.17 / 63.60 ± 1.63

28.68 ± 1.99 / 56.42 ± 3.34

65.23 ± 0.49 / 16.56 ± 0.94

38.44 ± 0.81 / 53.38 ± 0.60

37.69 ± 1.00 / 51.24 ± 0.73

12.6.1

mistralai/Mixtral-8x7B-v0.1 (few-shot)

46703

32

32768

True

2,363 ± 794 / 311 ± 105

2.07

59.10 ± 1.24 / 46.00 ± 3.08

55.30 ± 4.28 / 67.32 ± 5.07

37.84 ± 4.60 / 67.45 ± 2.49

31.71 ± 1.34 / 59.85 ± 2.39

64.23 ± 1.65 / 18.85 ± 2.99

47.89 ± 0.67 / 60.86 ± 0.50

28.53 ± 3.75 / 44.58 ± 2.73

14.0.4

google/gemma-7b (few-shot)

8538

256

8096

True

1,378 ± 260 / 387 ± 119

2.08

39.88 ± 2.56 / 35.40 ± 2.63

56.23 ± 3.17 / 68.87 ± 2.73

32.71 ± 1.60 / 64.55 ± 1.54

36.58 ± 2.81 / 64.92 ± 2.97

69.41 ± 1.20 / 29.96 ± 3.00

41.56 ± 0.82 / 54.79 ± 0.60

30.77 ± 2.85 / 43.20 ± 2.53

12.9.1

12.10.0

utter-project/EuroLLM-9B-Instruct (few-shot)

9152

128

4096

False

1,483 ± 321 / 379 ± 158

2.08

40.85 ± 2.21 / 34.64 ± 2.00

56.53 ± 2.49 / 69.48 ± 1.70

24.74 ± 4.45 / 56.60 ± 5.56

38.20 ± 1.47 / 67.94 ± 2.19

69.44 ± 1.03 / 28.76 ± 2.71

39.43 ± 1.06 / 54.28 ± 0.77

37.67 ± 1.13 / 51.67 ± 1.11

13.1.0

senseable/WestLake-7B-v2 (few-shot)

7242

32

32768

False

5,993 ± 1,028 / 1,742 ± 561

2.09

64.38 ± 1.60 / 50.26 ± 2.53

54.44 ± 1.45 / 69.32 ± 1.02

26.03 ± 2.23 / 61.88 ± 1.38

25.68 ± 2.81 / 62.48 ± 2.93

68.16 ± 0.95 / 24.52 ± 2.45

33.84 ± 1.54 / 50.24 ± 1.19

50.99 ± 0.99 / 63.11 ± 0.75

12.6.1

VAGOsolutions/FC-SauerkrautLM-7b-beta (few-shot)

unknown

32

8192

False

2,160 ± 514 / 668 ± 256

2.10

56.70 ± 1.55 / 41.89 ± 2.00

53.39 ± 1.89 / 67.16 ± 1.65

35.64 ± 1.50 / 66.83 ± 0.97

34.22 ± 1.43 / 67.00 ± 1.27

64.79 ± 0.22 / 17.18 ± 0.74

32.85 ± 0.79 / 49.55 ± 0.59

39.00 ± 2.60 / 53.77 ± 2.03

12.6.1

meta-llama/Meta-Llama-3-8B (few-shot)

8030

128

8192

True

1,477 ± 376 / 285 ± 97

2.10

56.00 ± 1.94 / 43.49 ± 2.05

56.40 ± 3.89 / 70.17 ± 2.91

22.01 ± 5.17 / 56.97 ± 3.54

35.39 ± 2.49 / 64.61 ± 2.42

68.92 ± 0.99 / 26.93 ± 2.01

38.12 ± 0.75 / 53.52 ± 0.56

31.37 ± 1.37 / 47.65 ± 1.09

12.6.1

mgoin/Nemotron-4-340B-Instruct-hf-FP8 (few-shot)

341029

256

4096

True

1,904 ± 475 / 361 ± 121

2.10

50.17 ± 7.37 / 32.80 ± 4.21

43.41 ± 7.58 / 60.96 ± 6.31

40.59 ± 9.15 / 63.58 ± 7.55

19.75 ± 2.91 / 43.00 ± 6.51

60.34 ± 1.55 / 10.97 ± 0.78

52.07 ± 8.24 / 62.36 ± 7.80

74.99 ± 6.86 / 80.52 ± 5.91

14.0.4

microsoft/Phi-3-mini-128k-instruct (few-shot)

3821

32

130976

True

7,312 ± 1,668 / 1,609 ± 525

2.10

55.36 ± 0.81 / 36.14 ± 1.96

53.05 ± 2.34 / 65.57 ± 1.92

23.08 ± 1.54 / 58.65 ± 2.04

31.55 ± 1.09 / 62.02 ± 2.17

67.33 ± 0.98 / 24.07 ± 2.44

39.57 ± 0.74 / 54.67 ± 0.56

51.26 ± 0.93 / 63.03 ± 0.69

12.9.1

12.10.0

CohereForAI/aya-expanse-8b (few-shot)

8028

256

8192

False

2,686 ± 685 / 491 ± 164

2.12

59.95 ± 1.43 / 39.14 ± 2.24

55.39 ± 1.97 / 69.86 ± 1.34

30.59 ± 1.76 / 64.21 ± 1.65

26.94 ± 1.06 / 57.13 ± 1.35

66.11 ± 0.37 / 18.21 ± 0.62

38.63 ± 1.03 / 54.04 ± 0.73

48.53 ± 1.17 / 61.25 ± 0.85

14.1.2

meta-llama/Llama-2-70b-chat-hf (few-shot, val)

68977

32

4125

True

1,979 ± 621 / 320 ± 105

2.12

62.39 ± 2.72 / 50.86 ± 2.31

53.16 ± 3.17 / 64.24 ± 3.42

31.81 ± 5.15 / 62.15 ± 4.02

28.99 ± 2.22 / 60.53 ± 2.92

66.98 ± 0.87 / 22.66 ± 2.41

35.72 ± 3.05 / 51.99 ± 2.09

35.26 ± 3.73 / 50.59 ± 3.17

12.7.0

meta-llama/Llama-3.1-8B-Instruct (few-shot)

8030

128

131072

True

1,473 ± 377 / 283 ± 96

2.12

62.03 ± 1.17 / 45.31 ± 1.89

58.15 ± 2.40 / 70.81 ± 1.86

30.18 ± 1.92 / 63.41 ± 1.19

26.48 ± 0.98 / 52.41 ± 1.48

64.68 ± 0.10 / 15.03 ± 0.29

39.36 ± 0.79 / 54.48 ± 0.57

47.21 ± 1.02 / 60.39 ± 0.77

14.1.2

occiglot/occiglot-7b-de-en-instruct (few-shot)

7242

32

32768

False

1,584 ± 217 / 635 ± 178

2.15

55.76 ± 1.16 / 40.04 ± 3.21

55.91 ± 2.49 / 70.31 ± 1.76

22.47 ± 3.37 / 56.77 ± 3.69

35.95 ± 1.89 / 66.86 ± 2.33

68.08 ± 0.49 / 25.46 ± 1.45

33.77 ± 0.86 / 49.44 ± 0.57

33.15 ± 1.86 / 48.23 ± 1.64

12.5.2

12.3.1

12.4.0

12.3.1

ibm-granite/granite-3.0-8b-instruct (few-shot)

8171

49

4096

True

1,118 ± 302 / 184 ± 63

2.21

54.68 ± 1.38 / 46.36 ± 2.67

55.48 ± 2.67 / 69.91 ± 1.90

26.89 ± 0.86 / 62.51 ± 0.48

31.27 ± 1.89 / 62.30 ± 2.09

66.33 ± 0.75 / 20.58 ± 1.77

33.98 ± 0.82 / 50.12 ± 0.65

34.59 ± 1.42 / 49.13 ± 1.25

13.0.0

nvidia/mistral-nemo-minitron-8b-base (few-shot)

8414

131

8192

True

2,470 ± 836 / 326 ± 111

2.22

59.82 ± 2.10 / 47.76 ± 2.63

50.22 ± 5.25 / 58.77 ± 6.76

30.43 ± 3.65 / 61.43 ± 3.32

30.22 ± 1.60 / 58.72 ± 2.48

62.56 ± 1.20 / 14.72 ± 2.03

43.09 ± 0.66 / 56.95 ± 0.53

33.40 ± 3.57 / 46.78 ± 3.16

14.1.1

mistralai/Mistral-7B-v0.1 (few-shot)

7242

32

32768

True

1,446 ± 354 / 295 ± 100

2.23

55.37 ± 1.32 / 44.65 ± 2.48

54.27 ± 1.71 / 68.13 ± 1.16

23.12 ± 4.07 / 57.81 ± 3.70

31.89 ± 3.29 / 59.77 ± 4.31

68.24 ± 0.70 / 25.71 ± 1.33

35.63 ± 1.12 / 51.69 ± 0.85

26.40 ± 1.86 / 43.98 ± 1.58

9.1.2

12.5.1

11.0.0

9.1.2

9.2.0

VAGOsolutions/SauerkrautLM-Gemma-7b (few-shot)

8538

256

8221

False

1,410 ± 299 / 370 ± 127

2.24

41.41 ± 2.64 / 26.93 ± 1.92

52.78 ± 4.59 / 66.87 ± 4.26

27.75 ± 1.14 / 63.54 ± 0.69

29.24 ± 2.44 / 58.33 ± 3.45

66.11 ± 1.41 / 22.22 ± 3.65

42.38 ± 0.74 / 56.56 ± 0.55

44.47 ± 2.47 / 56.41 ± 2.21

12.10.0

alpindale/Mistral-7B-v0.2-hf (few-shot)

7242

32

32768

True

1,841 ± 297 / 651 ± 193

2.24

55.32 ± 1.55 / 48.33 ± 1.45

52.49 ± 2.16 / 67.50 ± 1.61

24.34 ± 2.29 / 59.66 ± 2.93

31.54 ± 3.00 / 59.96 ± 3.89

68.98 ± 1.14 / 28.30 ± 2.36

35.12 ± 0.59 / 51.23 ± 0.46

28.89 ± 1.81 / 45.85 ± 1.47

12.5.2

mistralai/Mistral-7B-v0.3 (few-shot)

7248

33

32768

True

1,364 ± 343 / 266 ± 90

2.24

55.41 ± 1.45 / 48.39 ± 1.46

52.58 ± 2.42 / 67.52 ± 1.82

24.10 ± 2.12 / 59.47 ± 2.92

31.52 ± 2.95 / 60.03 ± 3.81

68.96 ± 1.13 / 28.26 ± 2.32

35.06 ± 0.54 / 51.20 ± 0.43

28.85 ± 1.70 / 45.83 ± 1.39

12.10.4

12.10.5

12.10.4

CohereForAI/c4ai-command-r-v01 (few-shot)

34981

256

8192

False

1,919 ± 645 / 248 ± 83

2.26

62.43 ± 0.91 / 43.18 ± 2.22

53.90 ± 1.12 / 65.63 ± 1.62

29.68 ± 3.64 / 62.39 ± 2.81

19.35 ± 3.19 / 50.25 ± 1.61

64.80 ± 0.12 / 14.58 ± 0.34

41.22 ± 0.72 / 55.63 ± 0.54

46.05 ± 1.33 / 59.27 ± 1.10

14.0.4

14.1.1

Nexusflow/Starling-LM-7B-beta (few-shot)

7242

32

4096

False

4,136 ± 1,282 / 668 ± 326

2.27

61.37 ± 1.04 / 36.06 ± 2.27

51.38 ± 2.17 / 67.66 ± 1.45

35.58 ± 1.85 / 66.99 ± 0.83

19.92 ± 3.88 / 57.70 ± 2.99

63.64 ± 0.06 / 12.60 ± 0.12

33.87 ± 0.63 / 49.89 ± 0.56

48.46 ± 1.55 / 60.82 ± 1.32

14.1.2

occiglot/occiglot-7b-eu5-instruct (few-shot)

7242

32

32768

False

2,088 ± 352 / 706 ± 214

2.27

52.63 ± 1.89 / 42.99 ± 2.40

43.16 ± 4.45 / 57.79 ± 4.61

27.09 ± 1.92 / 60.29 ± 1.99

34.01 ± 4.01 / 63.29 ± 3.97

69.43 ± 0.97 / 29.48 ± 2.59

32.56 ± 1.16 / 49.02 ± 0.79

23.61 ± 2.42 / 41.55 ± 2.14

12.5.2

12.2.0

12.3.1

12.4.0

12.3.1

occiglot/occiglot-7b-de-en (few-shot)

7242

32

32768

True

1,992 ± 319 / 706 ± 211

2.28

48.11 ± 2.01 / 39.66 ± 3.29

54.96 ± 2.69 / 69.64 ± 2.09

21.57 ± 4.18 / 55.63 ± 4.56

31.49 ± 3.11 / 61.33 ± 3.41

68.88 ± 0.67 / 27.13 ± 1.71

32.39 ± 0.86 / 49.12 ± 0.62

29.84 ± 1.54 / 46.53 ± 1.31

12.3.2

12.3.1

meta-llama/Llama-2-13b-hf (few-shot)

13016

32

4000

True

2,898 ± 637 / 736 ± 236

2.32

52.08 ± 1.86 / 46.67 ± 1.17

46.38 ± 2.70 / 60.32 ± 2.80

22.39 ± 3.96 / 57.26 ± 4.22

33.43 ± 2.38 / 62.65 ± 2.83

69.50 ± 1.29 / 29.74 ± 3.10

28.79 ± 1.53 / 45.87 ± 1.26

21.50 ± 2.52 / 39.80 ± 2.47

12.10.5

12.10.4

12.10.5

12.10.4

seedboxai/KafkaLM-7B-German-V0.1-DPO (few-shot)

7242

32

4096

False

6,070 ± 1,042 / 1,769 ± 573

2.32

48.92 ± 2.76 / 38.62 ± 2.42

52.57 ± 1.74 / 61.25 ± 2.84

20.74 ± 3.20 / 56.59 ± 3.27

32.87 ± 1.83 / 62.31 ± 2.13

68.88 ± 0.80 / 27.84 ± 2.07

31.38 ± 0.94 / 48.36 ± 0.62

29.83 ± 1.76 / 46.77 ± 1.39

12.5.2

12.3.2

12.4.0

12.3.2

meta-llama/Llama-3.1-8B (few-shot)

8030

128

131072

True

2,986 ± 823 / 276 ± 94

2.34

59.45 ± 1.64 / 46.60 ± 2.02

53.39 ± 5.35 / 65.74 ± 5.74

23.87 ± 5.74 / 57.17 ± 6.01

27.14 ± 2.19 / 54.63 ± 2.74

64.01 ± 1.18 / 17.85 ± 1.99

36.56 ± 0.73 / 52.11 ± 0.56

30.05 ± 2.57 / 46.82 ± 2.11

14.1.2

occiglot/occiglot-7b-eu5 (few-shot)

7242

32

32768

True

2,219 ± 427 / 717 ± 224

2.34

51.39 ± 1.35 / 44.47 ± 2.77

47.30 ± 4.44 / 62.28 ± 4.24

21.83 ± 1.98 / 57.05 ± 2.18

31.55 ± 3.67 / 60.39 ± 4.29

69.31 ± 0.68 / 28.27 ± 1.70

32.49 ± 0.91 / 48.82 ± 0.70

22.25 ± 2.13 / 40.28 ± 1.67

12.5.2

12.1.0

12.2.0

seedboxai/KafkaLM-7B-German-V0.1 (few-shot)

7242

32

32768

True

6,065 ± 1,038 / 1,766 ± 570

2.34

48.35 ± 2.96 / 38.58 ± 2.35

52.51 ± 1.72 / 61.27 ± 2.77

20.36 ± 3.59 / 56.14 ± 3.61

32.88 ± 1.78 / 62.19 ± 2.05

68.82 ± 0.81 / 27.89 ± 2.09

31.36 ± 0.96 / 48.34 ± 0.63

29.98 ± 1.64 / 46.82 ± 1.33

12.5.2

12.3.2

neuralmagic/Sparse-Llama-3.1-8B-2of4 (few-shot)

8030

128

131072

True

2,996 ± 817 / 284 ± 96

2.35

40.63 ± 2.51 / 38.86 ± 1.98

50.44 ± 2.55 / 62.82 ± 2.61

10.79 ± 1.67 / 40.21 ± 2.83

34.67 ± 1.48 / 61.01 ± 1.78

69.06 ± 1.32 / 29.21 ± 3.24

31.77 ± 0.71 / 48.72 ± 0.47

41.12 ± 1.38 / 53.80 ± 1.47

13.2.0

claude-3-5-sonnet-20241022 (zero-shot, val)

unknown

200000

True

193 ± 87 / 55 ± 19

2.36

61.83 ± 1.50 / 46.40 ± 1.54

61.59 ± 4.13 / 73.65 ± 2.62

46.40 ± 3.16 / 69.51 ± 1.58

23.77 ± 1.57 / 48.70 ± 1.26

62.69 ± 0.14 / 10.95 ± 0.17

17.15 ± 1.59 / 27.23 ± 2.17

9.70 ± 2.37 / 29.80 ± 1.51

14.0.3

meta-llama/Llama-2-13b-chat-hf (few-shot)

13016

32

4096

True

2,849 ± 622 / 723 ± 229

2.36

57.57 ± 1.46 / 44.84 ± 2.12

49.40 ± 2.11 / 63.02 ± 2.68

23.32 ± 2.20 / 59.16 ± 2.36

30.24 ± 1.18 / 61.10 ± 1.87

68.01 ± 1.28 / 25.44 ± 3.23

27.06 ± 0.92 / 45.06 ± 0.75

26.96 ± 0.49 / 44.79 ± 0.39

12.11.0

12.10.4

12.11.0

12.10.4

NorwAI/NorwAI-Mixtral-8x7B-instruct (few-shot)

46998

68

32768

True

9,015 ± 2,966 / 1,121 ± 510

2.37

56.14 ± 1.69 / 44.49 ± 3.25

53.33 ± 4.66 / 64.05 ± 6.18

29.49 ± 6.46 / 59.01 ± 6.80

18.49 ± 4.03 / 55.32 ± 2.95

64.84 ± 1.06 / 21.23 ± 1.72

44.31 ± 0.88 / 57.84 ± 0.67

30.72 ± 4.07 / 46.97 ± 3.37

14.0.4

ibm-granite/granite-3.0-8b-base (few-shot)

8171

49

4096

True

2,515 ± 625 / 476 ± 159

2.38

50.43 ± 1.80 / 41.26 ± 3.08

57.84 ± 3.59 / 71.20 ± 2.77

22.58 ± 5.59 / 53.17 ± 6.23

27.96 ± 1.39 / 54.99 ± 1.75

65.08 ± 1.62 / 20.35 ± 3.02

35.21 ± 1.13 / 51.21 ± 0.96

26.48 ± 1.04 / 43.71 ± 0.74

14.1.2

seedboxai/KafkaLM-13B-German-V0.1-DPO (few-shot)

13016

32

8221

False

789 ± 78 / 346 ± 102

2.45

52.45 ± 2.98 / 47.91 ± 2.37

49.73 ± 2.78 / 63.68 ± 2.80

20.72 ± 4.25 / 53.77 ± 5.69

30.18 ± 4.61 / 60.07 ± 5.86

68.97 ± 0.69 / 26.66 ± 1.82

27.71 ± 0.66 / 45.34 ± 0.63

21.44 ± 2.29 / 40.21 ± 1.74

12.7.0

12.10.4

12.7.0

12.10.0

NorwAI/NorwAI-Mixtral-8x7B (few-shot)

46998

68

32768

True

2,368 ± 793 / 317 ± 108

2.46

57.26 ± 1.41 / 42.54 ± 2.77

54.57 ± 4.98 / 68.40 ± 4.08

26.52 ± 4.63 / 59.83 ± 5.12

19.96 ± 1.50 / 43.10 ± 2.18

61.10 ± 0.91 / 14.12 ± 1.59

43.30 ± 1.06 / 57.22 ± 0.87

23.82 ± 3.21 / 41.16 ± 2.65

14.0.4

mistralai/Mistral-7B-Instruct-v0.2 (few-shot)

7242

32

32768

False

2,370 ± 416 / 711 ± 242

2.46

55.10 ± 1.35 / 41.89 ± 1.61

47.69 ± 2.35 / 64.93 ± 1.71

24.14 ± 2.09 / 60.83 ± 1.63

23.93 ± 2.11 / 57.64 ± 1.89

67.51 ± 0.71 / 22.63 ± 1.73

26.06 ± 1.65 / 44.13 ± 1.29

31.09 ± 1.35 / 47.48 ± 0.98

12.9.0

RuterNorway/Llama-2-13b-chat-norwegian (few-shot)

unknown

32

4125

False

3,254 ± 1,068 / 484 ± 173

2.47

57.02 ± 1.39 / 47.95 ± 2.09

49.75 ± 2.02 / 62.41 ± 3.30

19.80 ± 3.22 / 52.76 ± 5.45

27.86 ± 2.01 / 57.65 ± 2.01

66.83 ± 0.91 / 23.62 ± 2.07

25.99 ± 1.00 / 44.39 ± 0.75

24.90 ± 1.40 / 43.26 ± 1.22

12.9.0

nvidia/mistral-nemo-minitron-8b-instruct (few-shot)

8414

131

8192

True

3,161 ± 676 / 1,247 ± 481

2.49

60.69 ± 1.19 / 45.82 ± 1.42

53.77 ± 2.21 / 66.24 ± 2.25

38.53 ± 2.04 / 65.65 ± 2.18

0.00 ± 0.00 / 32.27 ± 3.90

62.73 ± 0.12 / 9.78 ± 0.11

45.91 ± 0.75 / 59.44 ± 0.55

47.37 ± 2.99 / 59.38 ± 2.38

14.1.1

timpal0l/Mistral-7B-v0.1-flashback-v2 (few-shot)

7242

32

32768

True

5,054 ± 1,200 / 1,056 ± 339

2.52

50.66 ± 1.53 / 39.89 ± 2.43

54.79 ± 3.53 / 68.79 ± 3.00

20.17 ± 1.69 / 58.67 ± 1.13

27.86 ± 4.70 / 54.38 ± 5.91

65.53 ± 1.07 / 19.46 ± 1.48

27.04 ± 1.04 / 44.99 ± 0.77

17.47 ± 3.26 / 36.70 ± 2.93

12.5.3

google/gemma-2-2b-it (few-shot)

2614

256

8224

True

5,374 ± 1,233 / 1,193 ± 377

2.54

37.31 ± 2.28 / 31.09 ± 2.10

46.23 ± 2.32 / 63.45 ± 1.57

23.26 ± 1.16 / 60.73 ± 0.24

28.01 ± 1.95 / 59.66 ± 2.10

66.35 ± 0.99 / 21.43 ± 2.27

30.11 ± 1.02 / 47.48 ± 0.79

35.25 ± 0.64 / 50.80 ± 0.55

13.0.0

mistralai/Mistral-7B-Instruct-v0.1 (few-shot)

7242

32

32797

False

634 ± 179 / 110 ± 35

2.55

47.19 ± 3.74 / 33.02 ± 2.09

47.26 ± 3.14 / 63.48 ± 2.94

22.32 ± 1.78 / 56.73 ± 4.00

24.36 ± 3.78 / 54.61 ± 4.44

67.75 ± 1.10 / 25.91 ± 2.95

26.79 ± 1.01 / 44.58 ± 0.85

20.33 ± 1.63 / 39.63 ± 1.09

12.9.0

CohereForAI/aya-23-8B (few-shot)

8028

256

8192

False

2,707 ± 688 / 497 ± 166

2.56

51.20 ± 4.11 / 41.29 ± 2.80

47.79 ± 3.34 / 64.03 ± 2.90

18.04 ± 1.40 / 54.89 ± 2.23

29.46 ± 1.12 / 55.85 ± 1.72

62.58 ± 0.45 / 15.06 ± 0.66

24.21 ± 0.88 / 42.71 ± 0.85

30.55 ± 3.52 / 45.97 ± 3.24

14.1.2

01-ai/Yi-1.5-6B (few-shot)

6061

64

4128

True

2,867 ± 550 / 793 ± 253

2.57

48.20 ± 1.57 / 39.04 ± 2.38

47.12 ± 4.20 / 59.69 ± 4.45

12.39 ± 1.39 / 45.13 ± 3.93

30.50 ± 3.67 / 57.48 ± 4.51

65.48 ± 1.73 / 22.72 ± 3.33

30.83 ± 1.40 / 48.12 ± 1.03

24.90 ± 1.67 / 43.21 ± 1.49

13.0.0

DiscoResearch/DiscoLM_German_7b_v1 (few-shot)

7242

32

32768

False

1,972 ± 315 / 689 ± 204

2.58

42.39 ± 2.43 / 32.42 ± 1.53

48.67 ± 3.85 / 59.21 ± 4.18

8.72 ± 2.15 / 43.37 ± 3.69

36.12 ± 2.35 / 66.54 ± 2.34

68.47 ± 1.27 / 26.03 ± 3.68

23.99 ± 1.15 / 41.69 ± 0.96

20.76 ± 1.28 / 38.13 ± 1.29

12.5.2

12.0.0

12.1.0

12.4.0

12.1.0

meta-llama/Llama-3.2-3B-Instruct (few-shot)

3213

128

131104

False

10,424 ± 2,641 / 2,081 ± 666

2.58

55.52 ± 2.07 / 46.18 ± 2.32

50.52 ± 2.29 / 62.39 ± 2.63

9.87 ± 2.95 / 42.20 ± 3.60

20.20 ± 3.28 / 47.02 ± 5.20

66.40 ± 0.98 / 21.21 ± 2.33

33.58 ± 0.41 / 50.27 ± 0.28

28.97 ± 1.01 / 46.45 ± 0.80

13.0.0

meta-llama/Llama-3.2-3B (few-shot)

3213

128

131104

True

3,713 ± 877 / 836 ± 267

2.58

49.85 ± 1.96 / 41.04 ± 2.44

54.65 ± 1.58 / 65.94 ± 2.42

3.17 ± 5.20 / 36.54 ± 5.71

29.37 ± 3.48 / 58.09 ± 4.16

67.96 ± 1.38 / 25.30 ± 2.54

29.56 ± 0.63 / 46.21 ± 0.55

16.15 ± 1.34 / 35.40 ± 1.27

13.0.0

ibm-granite/granite-8b-code-base-4k (few-shot)

8055

49

4096

True

2,313 ± 423 / 682 ± 210

2.60

59.07 ± 0.78 / 51.00 ± 2.06

49.75 ± 2.44 / 65.93 ± 1.66

14.71 ± 2.21 / 53.75 ± 3.07

29.45 ± 2.13 / 56.14 ± 2.53

67.78 ± 1.54 / 26.66 ± 3.40

16.39 ± 1.29 / 37.35 ± 0.95

10.68 ± 1.41 / 32.52 ± 1.35

13.0.0

ibm-granite/granite-3.0-2b-base (few-shot)

2534

49

4128

True

10,187 ± 2,363 / 2,204 ± 737

2.61

45.81 ± 1.27 / 39.33 ± 2.24

34.61 ± 3.47 / 44.53 ± 3.65

16.19 ± 3.33 / 54.43 ± 3.53

28.25 ± 4.10 / 57.52 ± 4.47

67.90 ± 0.72 / 26.38 ± 1.63

28.28 ± 0.88 / 46.06 ± 0.73

28.25 ± 2.23 / 45.60 ± 2.02

13.0.0

ibm-granite/granite-3.0-2b-instruct (few-shot)

2634

49

4128

True

10,194 ± 2,403 / 2,193 ± 731

2.62

53.30 ± 0.87 / 47.62 ± 1.30

44.95 ± 3.84 / 62.77 ± 2.82

18.67 ± 1.83 / 57.35 ± 3.01

28.10 ± 1.17 / 58.38 ± 1.45

65.95 ± 0.58 / 20.96 ± 1.39

23.84 ± 0.97 / 42.69 ± 0.79

16.86 ± 1.12 / 35.87 ± 1.30

13.0.0

meta-llama/Llama-2-7b-hf (few-shot)

6738

32

4125

True

930 ± 310 / 128 ± 43

2.68

43.02 ± 1.93 / 32.69 ± 1.98

50.21 ± 2.43 / 65.81 ± 1.82

15.79 ± 2.35 / 53.25 ± 4.45

28.57 ± 5.09 / 55.54 ± 6.14

68.68 ± 0.96 / 26.95 ± 1.81

18.38 ± 1.36 / 38.12 ± 1.19

8.45 ± 1.86 / 30.07 ± 1.49

12.8.0

12.9.0

DiscoResearch/Llama3-German-8B (few-shot)

8030

128

8192

True

2,383 ± 462 / 673 ± 210

2.69

54.34 ± 2.04 / 46.23 ± 2.24

58.32 ± 1.14 / 70.24 ± 0.99

25.70 ± 5.38 / 58.58 ± 4.44

0.50 ± 0.11 / 35.35 ± 1.23

62.61 ± 1.58 / 13.97 ± 1.64

38.63 ± 0.90 / 53.83 ± 0.67

33.51 ± 2.18 / 49.52 ± 1.89

13.0.0

NbAiLab/nb-llama-3.1-70B (few-shot)

70554

128

131072

True

1,220 ± 411 / 158 ± 53

2.70

67.63 ± 1.29 / 58.41 ± 2.59

55.79 ± 4.32 / 64.18 ± 5.71

24.45 ± 4.07 / 59.54 ± 4.12

1.17 ± 0.37 / 26.56 ± 2.63

55.51 ± 1.24 / 8.89 ± 0.79

44.20 ± 1.65 / 57.95 ± 1.26

31.71 ± 2.07 / 47.88 ± 1.81

14.0.4

meta-llama/Llama-2-7b-chat-hf (few-shot)

6738

32

4125

False

2,643 ± 455 / 800 ± 247

2.71

50.09 ± 1.33 / 38.59 ± 1.66

46.52 ± 2.85 / 63.64 ± 2.10

15.23 ± 1.71 / 55.08 ± 1.88

25.54 ± 3.58 / 56.07 ± 3.76

67.62 ± 0.89 / 23.52 ± 2.43

20.12 ± 1.13 / 39.48 ± 1.02

13.98 ± 1.56 / 34.07 ± 1.30

12.9.0

12.10.0

microsoft/Phi-3-mini-4k-instruct (few-shot)

3821

32

2047

True

8,681 ± 1,650 / 2,177 ± 717

2.71

59.14 ± 1.40 / 43.16 ± 2.75

51.72 ± 2.44 / 65.74 ± 2.10

33.78 ± 2.17 / 65.03 ± 1.74

24.82 ± 1.35 / 53.54 ± 2.38

41.45 ± 11.63 / 10.13 ± 2.62

38.78 ± 0.85 / 53.77 ± 0.65

22.32 ± 9.88 / 39.90 ± 7.49

14.1.2

Qwen/Qwen1.5-4B-Chat (few-shot)

3950

152

32768

False

4,347 ± 893 / 1,135 ± 365

2.72

42.08 ± 1.65 / 36.90 ± 2.00

41.52 ± 3.53 / 57.69 ± 3.35

12.78 ± 3.75 / 46.43 ± 5.48

29.35 ± 2.51 / 59.90 ± 2.80

65.56 ± 1.65 / 20.45 ± 2.68

23.76 ± 0.70 / 42.77 ± 0.56

20.92 ± 1.16 / 40.56 ± 0.86

12.5.2

10.0.1

12.1.0

12.5.2

12.1.0

ibm-granite/granite-3.0-3b-a800m-instruct (few-shot)

3374

49

4096

True

10,246 ± 3,021 / 1,629 ± 550

2.72

47.31 ± 1.67 / 31.36 ± 2.00

48.28 ± 4.28 / 64.07 ± 3.81

14.08 ± 1.29 / 52.67 ± 2.70

28.37 ± 4.07 / 55.92 ± 4.76

61.97 ± 1.94 / 19.06 ± 1.18

22.99 ± 0.83 / 41.98 ± 0.66

20.06 ± 1.95 / 38.69 ± 1.78

13.0.0

ibm-granite/granite-8b-code-instruct-4k (few-shot)

8055

49

4096

True

5,617 ± 995 / 1,623 ± 540

2.73

54.45 ± 1.17 / 42.36 ± 2.59

43.62 ± 3.18 / 59.82 ± 2.70

15.24 ± 1.87 / 55.49 ± 2.89

26.00 ± 2.28 / 51.82 ± 2.70

66.68 ± 1.30 / 24.80 ± 3.14

15.81 ± 1.07 / 36.18 ± 0.94

9.60 ± 1.12 / 31.31 ± 1.02

13.0.0

ibm-granite/granite-7b-instruct (few-shot)

6738

32

4096

True

3,136 ± 558 / 942 ± 290

2.83

39.38 ± 0.94 / 30.79 ± 1.08

49.83 ± 1.61 / 66.00 ± 1.36

4.03 ± 1.54 / 40.66 ± 3.14

30.20 ± 1.18 / 58.73 ± 1.73

64.38 ± 0.37 / 16.34 ± 0.85

20.76 ± 0.96 / 40.07 ± 0.71

20.66 ± 1.25 / 40.16 ± 1.09

13.2.0

13.0.0

13.2.0

allenai/OLMo-1.7-7B-hf (few-shot)

6888

50

4096

True

3,371 ± 876 / 561 ± 184

2.85

39.41 ± 2.30 / 36.17 ± 2.32

49.42 ± 4.33 / 61.57 ± 5.43

6.02 ± 2.53 / 46.41 ± 4.35

27.69 ± 4.39 / 54.92 ± 5.38

66.75 ± 0.68 / 23.16 ± 0.88

20.77 ± 0.92 / 39.29 ± 0.91

10.47 ± 1.51 / 31.09 ± 1.36

12.10.5

12.10.4

12.10.5

12.10.4

claude-3-5-haiku-20241022 (zero-shot, val)

unknown

200000

True

277 ± 77 / 70 ± 25

2.87

55.59 ± 1.42 / 31.16 ± 1.29

43.73 ± 2.53 / 61.91 ± 1.47

23.74 ± 3.03 / 45.75 ± 2.40

21.36 ± 1.42 / 50.29 ± 1.55

62.97 ± 0.16 / 11.18 ± 0.21

16.53 ± 2.79 / 28.83 ± 2.66

-1.78 ± 3.04 / 25.27 ± 2.64

14.0.3

14.0.2

14.0.3

google/gemma-2-2b (few-shot)

2614

256

8224

True

5,235 ± 1,226 / 1,154 ± 366

2.89

19.69 ± 4.71 / 19.53 ± 3.13

50.36 ± 2.83 / 60.36 ± 3.45

9.07 ± 2.41 / 44.45 ± 4.37

27.06 ± 4.74 / 53.36 ± 6.20

68.52 ± 1.47 / 26.96 ± 3.01

26.50 ± 0.89 / 44.19 ± 0.75

9.36 ± 1.35 / 30.60 ± 1.18

13.0.0

google/gemma-7b-it (few-shot)

8538

256

8221

False

1,792 ± 249 / 668 ± 203

2.89

54.20 ± 1.00 / 48.58 ± 1.53

15.43 ± 3.70 / 43.11 ± 2.77

17.49 ± 1.23 / 57.46 ± 1.12

28.68 ± 5.36 / 60.07 ± 4.46

64.87 ± 0.47 / 22.25 ± 1.89

22.10 ± 0.68 / 40.39 ± 0.63

18.58 ± 1.48 / 36.49 ± 1.24

12.10.0

TrustLLMeu/baseline-7-8b_1t-tokens_llama (few-shot)

7800

100

4096

True

6,197 ± 1,118 / 1,730 ± 577

2.92

39.21 ± 2.29 / 36.08 ± 2.06

58.36 ± 1.80 / 71.98 ± 1.17

7.03 ± 3.09 / 50.18 ± 3.74

27.02 ± 3.09 / 51.94 ± 4.17

68.59 ± 0.76 / 25.49 ± 2.17

3.52 ± 0.84 / 26.38 ± 0.62

0.06 ± 0.82 / 24.56 ± 0.78

13.0.0

ibm-granite/granite-3.0-3b-a800m-base (few-shot)

3374

49

4096

True

10,504 ± 3,028 / 1,678 ± 559

2.92

40.61 ± 2.18 / 28.49 ± 2.11

31.86 ± 3.60 / 42.96 ± 3.98

5.36 ± 3.96 / 37.83 ± 4.03

25.99 ± 3.85 / 47.72 ± 4.74

66.77 ± 0.94 / 23.71 ± 1.90

22.17 ± 1.05 / 41.55 ± 0.85

22.61 ± 1.75 / 41.16 ± 1.46

13.0.0

Qwen/Qwen1.5-4B (few-shot)

3950

152

32768

True

3,248 ± 739 / 761 ± 252

2.95

31.52 ± 2.96 / 29.20 ± 1.88

39.91 ± 3.29 / 53.66 ± 3.20

3.27 ± 2.51 / 34.30 ± 1.29

27.55 ± 3.12 / 57.60 ± 3.34

65.88 ± 1.25 / 21.37 ± 1.87

21.32 ± 1.14 / 40.66 ± 0.96

21.35 ± 0.89 / 40.77 ± 0.65

12.5.2

10.0.1

12.1.0

01-ai/Yi-6B (few-shot)

6061

64

4125

True

6,435 ± 1,316 / 1,632 ± 549

2.96

0.00 ± 0.00 / 0.00 ± 0.00

52.66 ± 2.45 / 67.63 ± 1.87

7.33 ± 2.53 / 37.69 ± 2.51

30.05 ± 1.59 / 57.13 ± 2.48

66.09 ± 1.28 / 23.20 ± 2.54

30.33 ± 0.78 / 47.68 ± 0.54

22.89 ± 1.14 / 41.15 ± 1.08

12.10.0

ibm-granite/granite-7b-base (few-shot)

6738

32

2048

True

4,405 ± 1,098 / 1,032 ± 345

3.02

37.68 ± 1.26 / 33.74 ± 1.78

46.00 ± 3.63 / 61.88 ± 2.94

0.83 ± 0.84 / 33.38 ± 0.28

26.65 ± 4.01 / 53.15 ± 4.26

64.14 ± 0.38 / 18.78 ± 0.63

18.12 ± 0.96 / 37.66 ± 1.06

9.92 ± 2.59 / 31.17 ± 1.75

12.10.5

12.10.8

stabilityai/stablelm-2-1_6b (few-shot)

1645

100

4096

True

7,259 ± 2,120 / 1,240 ± 432

3.09

34.81 ± 2.51 / 30.33 ± 2.95

51.01 ± 2.18 / 65.35 ± 2.23

0.00 ± 0.00 / 33.34 ± 0.31

25.40 ± 3.91 / 49.38 ± 4.65

63.53 ± 2.45 / 17.65 ± 2.20

11.23 ± 0.71 / 33.23 ± 0.76

7.25 ± 1.63 / 29.15 ± 1.35

12.10.8

ibm-granite/granite-3b-code-instruct-2k (few-shot)

3483

49

2048

True

9,059 ± 1,947 / 2,201 ± 728

3.16

49.16 ± 2.12 / 40.34 ± 2.81

35.17 ± 4.48 / 51.15 ± 4.55

9.79 ± 2.06 / 50.77 ± 3.64

22.48 ± 2.59 / 45.90 ± 3.03

60.81 ± 0.38 / 15.92 ± 0.54

6.89 ± 0.76 / 28.82 ± 0.69

4.79 ± 0.91 / 28.40 ± 0.78

13.0.0

openGPT-X/Teuken-7B-instruct-research-v0.4 (few-shot)

7453

251

4096

False

1,254 ± 328 / 243 ± 83

3.18

39.39 ± 2.50 / 23.55 ± 1.34

23.60 ± 3.12 / 34.73 ± 3.27

7.68 ± 1.29 / 36.93 ± 0.75

25.30 ± 0.79 / 52.63 ± 1.81

63.84 ± 0.15 / 12.84 ± 0.30

14.99 ± 1.10 / 33.86 ± 1.26

14.99 ± 1.43 / 33.86 ± 1.03

14.0.4

Rijgersberg/GEITje-7B (few-shot)

7242

32

32797

True

5,887 ± 1,087 / 1,600 ± 522

3.19

0.00 ± 0.00 / 0.00 ± 0.00

47.67 ± 2.82 / 60.09 ± 3.28

9.67 ± 2.79 / 46.35 ± 4.48

26.23 ± 3.79 / 53.16 ± 4.52

65.42 ± 0.61 / 18.57 ± 0.70

19.03 ± 1.34 / 36.86 ± 1.84

9.87 ± 1.98 / 30.31 ± 1.95

12.10.0

meta-llama/Llama-3.2-1B-Instruct (few-shot)

1236

128

131104

False

7,436 ± 1,846 / 1,508 ± 479

3.23

42.51 ± 1.31 / 37.05 ± 1.59

38.26 ± 2.34 / 55.52 ± 2.66

5.48 ± 1.36 / 48.23 ± 2.44

19.43 ± 3.10 / 41.82 ± 3.29

60.81 ± 1.75 / 13.79 ± 2.22

16.06 ± 1.35 / 36.96 ± 0.92

6.64 ± 0.55 / 29.12 ± 0.60

13.0.0

VAGOsolutions/SauerkrautLM-Gemma-2b (few-shot)

2506

256

8192

False

3,607 ± 565 / 1,212 ± 349

3.29

12.21 ± 2.76 / 11.93 ± 2.08

44.84 ± 2.70 / 57.27 ± 3.65

2.02 ± 2.19 / 37.47 ± 3.26

24.59 ± 2.70 / 49.79 ± 2.96

62.87 ± 2.44 / 18.13 ± 1.20

16.35 ± 0.62 / 36.72 ± 0.48

8.93 ± 0.99 / 30.43 ± 0.75

12.6.1

utter-project/EuroLLM-1.7B-Instruct (few-shot)

1657

128

4096

True

15,009 ± 4,072 / 2,702 ± 878

3.30

28.49 ± 2.30 / 24.73 ± 1.76

43.18 ± 2.89 / 58.22 ± 3.50

2.92 ± 1.40 / 44.65 ± 3.51

23.26 ± 3.37 / 48.47 ± 3.57

64.64 ± 0.93 / 17.38 ± 1.77

2.60 ± 0.88 / 26.76 ± 0.79

1.46 ± 1.09 / 26.00 ± 0.95

13.1.0

google/gemma-2b (few-shot)

2506

256

8192

True

6,087 ± 1,046 / 1,902 ± 563

3.32

16.95 ± 2.96 / 15.80 ± 2.16

44.96 ± 3.30 / 61.27 ± 2.88

0.77 ± 1.22 / 33.68 ± 0.59

17.92 ± 4.72 / 40.68 ± 6.34

66.80 ± 0.96 / 22.24 ± 1.37

12.11 ± 0.94 / 33.12 ± 0.90

7.32 ± 1.40 / 30.11 ± 1.16

12.5.2

12.1.0

MaLA-LM/emma-500-llama2-7b (few-shot)

6738

32

4096

True

6,275 ± 1,193 / 1,755 ± 578

3.33

32.33 ± 2.48 / 30.20 ± 1.92

26.39 ± 5.23 / 36.06 ± 6.62

1.44 ± 1.38 / 33.60 ± 0.42

28.15 ± 5.57 / 54.13 ± 6.75

58.62 ± 3.80 / 13.54 ± 2.20

14.94 ± 1.09 / 35.44 ± 0.89

8.70 ± 0.96 / 31.09 ± 0.58

13.0.0

google/gemma-2b-it (few-shot)

2506

256

8192

False

6,471 ± 1,142 / 1,961 ± 584

3.33

36.62 ± 1.56 / 28.22 ± 1.66

28.54 ± 2.70 / 50.10 ± 1.65

1.15 ± 1.66 / 38.16 ± 2.78

23.39 ± 1.00 / 51.61 ± 1.04

63.02 ± 2.00 / 19.54 ± 1.33

12.27 ± 1.33 / 33.40 ± 1.08

6.57 ± 0.74 / 28.60 ± 0.70

12.5.2

12.1.0

12.4.0

12.1.0

openGPT-X/Teuken-7B-instruct-commercial-v0.4 (few-shot)

7453

251

4096

True

1,438 ± 410 / 233 ± 79

3.37

38.81 ± 2.72 / 27.14 ± 1.94

10.59 ± 3.02 / 21.22 ± 2.19

0.91 ± 1.67 / 33.45 ± 0.28

22.54 ± 1.17 / 48.13 ± 1.67

64.28 ± 0.13 / 14.60 ± 0.27

15.69 ± 0.87 / 35.90 ± 0.62

14.18 ± 2.07 / 33.42 ± 2.12

14.0.4

Qwen/Qwen1.5-1.8B-Chat (few-shot)

1837

152

32768

False

8,304 ± 1,846 / 1,933 ± 617

3.38

28.04 ± 2.71 / 24.08 ± 1.58

36.21 ± 3.42 / 54.82 ± 3.32

3.12 ± 1.42 / 46.21 ± 2.93

16.33 ± 3.22 / 41.91 ± 4.34

61.47 ± 1.67 / 12.62 ± 1.40

13.44 ± 1.22 / 34.26 ± 0.98

8.31 ± 1.08 / 31.05 ± 0.84

12.5.2

11.0.0

12.1.0

12.5.0

12.1.0

HuggingFaceTB/SmolLM2-1.7B-Instruct (few-shot)

1711

49

8192

True

15,971 ± 3,654 / 3,609 ± 1,197

3.41

32.54 ± 1.57 / 26.92 ± 1.87

27.03 ± 2.54 / 42.80 ± 3.28

8.95 ± 1.62 / 51.86 ± 2.23

18.38 ± 1.26 / 39.18 ± 2.13

59.91 ± 0.84 / 16.05 ± 0.61

12.30 ± 1.16 / 32.60 ± 1.27

6.28 ± 1.36 / 27.82 ± 1.17

13.1.0

ibm-granite/granite-3.0-1b-a400m-instruct (few-shot)

1335

49

4096

True

7,964 ± 2,255 / 1,299 ± 433

3.44

36.56 ± 2.08 / 25.71 ± 1.73

40.65 ± 2.11 / 58.49 ± 1.94

4.68 ± 1.88 / 45.72 ± 3.59

19.45 ± 0.59 / 39.53 ± 0.90

57.01 ± 0.95 / 10.04 ± 0.58

4.68 ± 1.02 / 27.76 ± 0.87

0.48 ± 0.63 / 24.59 ± 0.59

13.2.0

meta-llama/Llama-3.2-1B (few-shot)

1236

128

131104

True

7,577 ± 1,884 / 1,555 ± 492

3.45

24.79 ± 6.48 / 22.92 ± 5.74

47.65 ± 2.85 / 63.11 ± 2.17

2.39 ± 1.46 / 39.92 ± 4.38

13.39 ± 4.13 / 33.76 ± 5.50

61.07 ± 4.19 / 16.72 ± 4.33

3.94 ± 1.35 / 27.78 ± 1.07

0.90 ± 0.98 / 25.40 ± 0.65

13.0.0

seedboxai/KafkaLM-70B-German-V0.1 (few-shot, val)

68977

32

4125

True

294 ± 21 / 168 ± 42

3.45

59.23 ± 2.95 / 52.06 ± 3.03

-5.01 ± 4.03 / 18.00 ± 0.71

3.19 ± 4.99 / 37.63 ± 3.16

19.84 ± 2.17 / 56.60 ± 3.00

62.42 ± 4.28 / 22.39 ± 3.89

18.40 ± 3.05 / 34.38 ± 2.84

1.52 ± 1.25 / 26.80 ± 1.34

12.10.0

HuggingFaceTB/SmolLM2-1.7B (few-shot)

1711

49

8192

True

16,249 ± 3,690 / 3,689 ± 1,226

3.46

28.67 ± 3.31 / 25.27 ± 2.68

19.69 ± 2.50 / 29.00 ± 2.22

5.07 ± 0.89 / 47.60 ± 2.39

18.43 ± 2.31 / 38.33 ± 2.91

64.25 ± 1.99 / 18.90 ± 2.50

15.15 ± 0.67 / 35.75 ± 0.46

6.23 ± 0.96 / 28.39 ± 0.87

13.1.0

LumiOpen/Viking-13B (few-shot)

14030

131

4128

True

840 ± 79 / 400 ± 124

3.46

34.53 ± 1.24 / 29.89 ± 1.96

42.90 ± 2.66 / 56.64 ± 4.71

1.51 ± 1.64 / 43.36 ± 4.05

15.83 ± 1.42 / 29.77 ± 2.54

61.40 ± 3.15 / 14.84 ± 2.07

-1.84 ± 1.09 / 23.98 ± 0.66

-0.12 ± 1.30 / 24.41 ± 0.96

12.5.2

ibm-granite/granite-3b-code-base-2k (few-shot)

3483

49

2048

True

2,732 ± 868 / 662 ± 238

3.49

49.38 ± 2.20 / 42.66 ± 3.24

41.72 ± 4.07 / 60.45 ± 3.07

7.67 ± 1.52 / 46.66 ± 3.23

13.70 ± 2.43 / 30.21 ± 3.80

45.88 ± 3.36 / 8.30 ± 1.98

8.73 ± 1.03 / 31.46 ± 0.70

6.18 ± 1.33 / 29.78 ± 1.02

13.0.0

NorwAI/NorwAI-Llama2-7B (few-shot)

7033

68

4096

True

4,438 ± 1,128 / 1,028 ± 346

3.51

25.69 ± 1.43 / 25.95 ± 1.23

33.48 ± 2.83 / 47.14 ± 4.43

3.73 ± 1.14 / 44.43 ± 4.17

14.82 ± 2.69 / 35.50 ± 3.15

63.85 ± 2.30 / 21.16 ± 3.79

2.96 ± 0.44 / 26.82 ± 0.62

1.88 ± 0.93 / 26.35 ± 1.20

12.10.4

AI-Sweden-Models/gpt-sw3-20b-instruct (few-shot)

20918

64

2077

True

1,831 ± 587 / 268 ± 90

3.52

24.35 ± 1.72 / 21.90 ± 0.85

43.35 ± 3.81 / 60.49 ± 3.18

2.38 ± 1.21 / 37.27 ± 1.09

15.56 ± 2.24 / 34.68 ± 3.15

57.47 ± 0.69 / 14.10 ± 0.73

4.67 ± 0.91 / 27.88 ± 0.81

1.77 ± 1.29 / 25.47 ± 0.78

12.7.0

Qwen/Qwen1.5-1.8B (few-shot)

1837

152

32768

True

5,666 ± 1,328 / 1,256 ± 408

3.52

9.23 ± 4.86 / 10.43 ± 3.83

38.30 ± 2.90 / 56.94 ± 2.83

0.39 ± 1.17 / 33.47 ± 0.34

16.67 ± 3.02 / 41.61 ± 3.00

61.74 ± 1.58 / 14.71 ± 1.12

13.65 ± 1.17 / 33.35 ± 0.91

8.86 ± 0.99 / 30.21 ± 0.77

12.5.2

10.0.1

12.1.0

RuterNorway/Llama-2-7b-chat-norwegian (few-shot)

6738

32

4096

False

10,890 ± 2,686 / 2,186 ± 750

3.52

27.22 ± 1.38 / 24.48 ± 1.76

33.61 ± 5.06 / 49.68 ± 5.74

0.45 ± 0.91 / 35.24 ± 3.71

20.44 ± 3.29 / 45.50 ± 3.33

60.50 ± 0.63 / 13.71 ± 0.75

-0.10 ± 0.93 / 25.16 ± 1.17

-1.00 ± 1.03 / 24.94 ± 1.00

9.3.1

12.10.0

9.3.1

12.5.2

12.4.0

9.3.1

NorwAI/NorwAI-Mistral-7B (few-shot)

7537

68

4096

True

3,035 ± 503 / 911 ± 300

3.56

19.29 ± 5.77 / 20.20 ± 3.59

43.88 ± 3.42 / 61.04 ± 2.69

5.63 ± 1.32 / 49.95 ± 1.54

17.02 ± 2.10 / 36.47 ± 2.53

54.94 ± 1.06 / 14.92 ± 1.40

5.89 ± 0.78 / 27.89 ± 0.68

4.11 ± 0.97 / 27.05 ± 0.97

12.10.4

tiiuae/Falcon3-1B-Instruct (few-shot)

1669

131

8192

True

9,270 ± 2,690 / 1,434 ± 437

3.60

32.54 ± 1.23 / 30.63 ± 1.12

22.27 ± 4.97 / 36.09 ± 3.66

7.18 ± 1.13 / 51.77 ± 1.56

16.72 ± 0.88 / 38.07 ± 0.76

58.08 ± 0.37 / 10.60 ± 0.37

6.57 ± 1.03 / 29.10 ± 0.93

5.16 ± 1.64 / 28.12 ± 1.34

14.1.2

AI-Sweden-Models/gpt-sw3-20b (few-shot)

20918

64

2077

True

1,875 ± 673 / 261 ± 91

3.64

36.17 ± 2.52 / 27.29 ± 1.74

34.17 ± 7.08 / 46.97 ± 8.28

2.21 ± 1.64 / 38.29 ± 3.56

13.60 ± 3.04 / 30.89 ± 4.33

51.97 ± 2.69 / 10.46 ± 2.09

3.68 ± 1.19 / 27.54 ± 0.71

0.64 ± 1.58 / 25.05 ± 1.13

12.9.0

ibm-granite/granite-3.0-1b-a400m-base (few-shot)

1385

49

4096

True

7,808 ± 2,183 / 1,289 ± 428

3.65

19.94 ± 4.65 / 19.11 ± 3.71

34.66 ± 3.73 / 44.27 ± 4.29

0.70 ± 1.06 / 33.46 ± 0.38

14.66 ± 2.25 / 31.39 ± 2.73

61.26 ± 1.64 / 14.93 ± 1.37

-0.48 ± 1.78 / 22.64 ± 1.11

0.28 ± 0.88 / 25.12 ± 0.77

13.2.0

DiscoResearch/DiscoLM-70b (few-shot, val)

68977

32

8221

False

291 ± 23 / 163 ± 39

3.71

62.16 ± 2.83 / 48.46 ± 3.31

27.56 ± 3.62 / 47.04 ± 4.06

0.59 ± 1.16 / 32.41 ± 0.74

0.00 ± 0.00 / 4.44 ± 0.39

52.77 ± 0.13 / 4.84 ± 0.05

4.04 ± 2.64 / 22.46 ± 1.78

7.33 ± 5.87 / 28.83 ± 3.05

12.10.0

Qwen/Qwen1.5-0.5B (few-shot)

620

152

32768

True

11,371 ± 2,924 / 2,122 ± 692

3.81

27.34 ± 1.95 / 24.46 ± 1.25

10.64 ± 5.31 / 26.79 ± 4.73

0.33 ± 1.20 / 35.20 ± 2.45

11.81 ± 2.10 / 27.38 ± 2.49

59.71 ± 2.14 / 12.23 ± 1.08

6.34 ± 1.10 / 29.19 ± 0.75

2.94 ± 0.65 / 27.16 ± 0.61

12.5.2

10.0.1

12.1.0

PleIAs/Pleias-3b-Preview (few-shot)

3212

66

4096

True

6,513 ± 1,241 / 1,282 ± 644

3.91

23.08 ± 4.90 / 21.68 ± 4.10

7.41 ± 5.97 / 23.37 ± 5.40

0.89 ± 1.81 / 40.64 ± 3.86

17.32 ± 1.06 / 33.36 ± 2.38

56.57 ± 2.63 / 11.57 ± 1.39

-0.63 ± 0.62 / 23.00 ± 0.52

-0.17 ± 1.38 / 24.60 ± 0.84

14.0.4

14.1.2

14.0.4

Qwen/Qwen1.5-0.5B-Chat (few-shot)

620

152

32768

False

11,740 ± 3,000 / 2,209 ± 721

3.91

24.67 ± 0.99 / 23.98 ± 0.73

9.31 ± 2.97 / 21.50 ± 2.70

1.11 ± 1.69 / 37.88 ± 4.05

13.60 ± 1.60 / 29.10 ± 1.94

56.42 ± 7.64 / 11.68 ± 1.75

2.75 ± 0.91 / 27.17 ± 0.72

3.41 ± 1.30 / 27.45 ± 0.79

12.5.2

11.0.0

12.1.0

12.5.0

12.1.0

PleIAs/Pleias-1.2b-Preview (few-shot)

1195

66

2048

True

10,756 ± 3,589 / 1,157 ± 670

3.93

34.68 ± 3.56 / 30.00 ± 2.82

21.76 ± 5.01 / 39.73 ± 5.21

0.85 ± 1.86 / 38.42 ± 3.98

14.30 ± 0.97 / 30.11 ± 2.25

43.98 ± 8.28 / 9.06 ± 1.56

0.36 ± 1.05 / 25.80 ± 0.69

1.32 ± 0.86 / 25.98 ± 0.61

14.0.4

14.1.2

14.0.4

allenai/OLMo-7B (few-shot)

6888

50

2080

True

5,403 ± 1,133 / 1,294 ± 423

3.93

30.85 ± 4.69 / 24.38 ± 3.10

49.77 ± 2.81 / 64.87 ± 2.42

2.67 ± 1.77 / 41.55 ± 4.54

4.09 ± 1.94 / 12.70 ± 2.66

42.64 ± 0.22 / 4.37 ± 0.10

1.94 ± 0.76 / 26.86 ± 0.44

1.24 ± 0.75 / 25.78 ± 0.82

12.5.2

HuggingFaceTB/SmolLM2-360M (few-shot)

362

49

8192

True

22,023 ± 6,203 / 3,675 ± 1,231

3.98

19.94 ± 0.96 / 18.01 ± 0.59

19.64 ± 5.59 / 36.97 ± 5.41

0.00 ± 0.00 / 33.32 ± 0.30

8.78 ± 1.33 / 20.50 ± 1.50

57.06 ± 3.46 / 11.78 ± 1.69

0.32 ± 0.94 / 22.52 ± 0.61

-0.65 ± 0.72 / 24.04 ± 0.46

13.1.0

state-spaces/mamba-2.8b-hf (few-shot)

2768

50

32800

True

2,722 ± 495 / 766 ± 250

3.99

21.96 ± 1.53 / 18.48 ± 1.53

18.66 ± 3.01 / 35.11 ± 2.93

0.16 ± 1.78 / 37.84 ± 2.92

7.08 ± 1.91 / 18.42 ± 2.90

56.47 ± 2.05 / 11.00 ± 1.46

1.65 ± 1.53 / 26.62 ± 0.88

0.02 ± 1.24 / 25.61 ± 0.98

13.0.0

allenai/OLMo-7B-Twin-2T (few-shot)

6888

50

2080

True

5,484 ± 1,125 / 1,317 ± 425

4.17

14.06 ± 5.31 / 12.90 ± 4.52

28.07 ± 6.33 / 38.61 ± 7.42

2.31 ± 1.88 / 44.45 ± 4.23

6.89 ± 2.35 / 17.95 ± 3.37

43.42 ± 0.26 / 5.42 ± 0.11

1.66 ± 1.18 / 25.51 ± 0.86

1.50 ± 0.90 / 26.46 ± 0.95

12.5.2

HuggingFaceTB/SmolLM2-360M-Instruct (few-shot)

362

49

8192

True

21,777 ± 6,115 / 3,617 ± 1,211

4.21

18.77 ± 3.96 / 18.65 ± 3.26

12.59 ± 3.85 / 22.64 ± 2.25

1.64 ± 1.30 / 34.84 ± 2.12

9.27 ± 1.07 / 21.58 ± 1.37

46.31 ± 1.46 / 8.68 ± 0.61

0.84 ± 0.81 / 22.50 ± 0.61

0.01 ± 1.07 / 24.43 ± 0.50

13.1.0

PleIAs/Pleias-Nano (few-shot)

1195

66

2048

True

2,519 ± 841 / 323 ± 104

4.21

24.32 ± 2.91 / 22.08 ± 1.97

15.58 ± 3.73 / 29.45 ± 4.32

1.25 ± 1.70 / 37.57 ± 3.82

6.82 ± 2.39 / 20.51 ± 3.06

44.90 ± 8.88 / 8.06 ± 1.63

-0.19 ± 1.08 / 24.99 ± 1.08

1.18 ± 0.93 / 26.03 ± 0.59

14.0.4

14.1.2

14.0.4

PleIAs/Pleias-350m-Preview (few-shot)

353

66

2048

True

10,242 ± 3,432 / 1,335 ± 484

4.27

25.30 ± 2.32 / 24.21 ± 2.13

13.90 ± 5.43 / 30.42 ± 6.17

-0.25 ± 1.90 / 39.25 ± 4.50

6.12 ± 0.63 / 15.65 ± 1.66

41.90 ± 7.85 / 7.52 ± 1.26

-1.81 ± 1.16 / 22.39 ± 0.62

0.00 ± 0.84 / 24.71 ± 0.61

14.0.4

14.1.2

14.0.4

PleIAs/Pleias-Pico (few-shot)

353

66

2048

True

2,331 ± 787 / 301 ± 97

4.28

21.03 ± 2.96 / 21.02 ± 2.73

10.99 ± 6.72 / 27.88 ± 6.96

0.13 ± 1.96 / 36.48 ± 3.10

3.61 ± 0.37 / 11.45 ± 1.56

43.73 ± 8.31 / 7.34 ± 1.49

-0.77 ± 0.77 / 22.86 ± 0.58

0.11 ± 0.91 / 24.98 ± 0.79

14.0.4

14.1.2

14.0.4

HuggingFaceTB/SmolLM2-135M (few-shot)

135

49

8192

True

26,346 ± 7,812 / 4,082 ± 1,372

4.31

16.89 ± 1.62 / 16.63 ± 1.80

2.74 ± 3.46 / 23.30 ± 3.11

-0.34 ± 1.06 / 39.21 ± 4.13

0.28 ± 0.27 / 2.91 ± 0.92

54.79 ± 2.10 / 8.47 ± 1.10

-0.32 ± 0.88 / 25.45 ± 0.97

0.18 ± 0.99 / 25.72 ± 0.91

13.1.0

allenai/OLMo-1B (few-shot)

1177

50

2080

True

8,536 ± 1,926 / 1,940 ± 619

4.38

21.46 ± 2.04 / 20.83 ± 1.63

21.03 ± 6.33 / 38.33 ± 7.79

0.13 ± 1.48 / 43.17 ± 4.90

0.71 ± 0.53 / 6.02 ± 1.37

39.77 ± 0.20 / 4.80 ± 0.05

-0.08 ± 0.53 / 25.19 ± 1.04

-0.63 ± 0.73 / 24.55 ± 0.49

12.5.2

12.1.0

HuggingFaceTB/SmolLM2-135M-Instruct (few-shot)

135

49

8192

True

25,602 ± 7,583 / 3,953 ± 1,325

4.40

15.54 ± 1.74 / 15.29 ± 1.88

2.51 ± 1.90 / 21.27 ± 3.22

0.36 ± 1.07 / 39.04 ± 3.21

1.77 ± 0.67 / 5.69 ± 0.96

49.41 ± 1.73 / 7.27 ± 0.41

-1.46 ± 0.84 / 24.84 ± 0.68

-0.18 ± 1.25 / 24.79 ± 1.02

13.1.0

RJuro/kanelsnegl-v0.1 (few-shot)

7242

32

541

True

5,847 ± 1,029 / 1,640 ± 525

4.40

0.00 ± 0.00 / 0.00 ± 0.00

0.00 ± 0.00 / 17.05 ± 0.35

0.00 ± 0.00 / 33.34 ± 0.31

0.00 ± 0.00 / 14.17 ± 0.79

59.26 ± 0.09 / 9.32 ± 0.09

1.16 ± 0.57 / 22.46 ± 0.56

0.31 ± 0.50 / 24.18 ± 0.53

12.10.0

NorwAI/NorwAI-Mistral-7B-pretrain (few-shot)

7537

68

4096

True

3,024 ± 496 / 909 ± 301

4.75

5.80 ± 1.56 / 5.41 ± 1.56

4.45 ± 1.73 / 29.26 ± 3.66

-0.48 ± 1.33 / 43.09 ± 3.56

0.08 ± 0.12 / 4.02 ± 1.17

35.92 ± 0.71 / 3.53 ± 0.24

0.77 ± 0.60 / 23.33 ± 0.60

-0.34 ± 1.16 / 24.39 ± 0.86

12.10.4

ssmits/Falcon2-5.5B-multilingual (few-shot)

5465

65

8192

True

7,692 ± 1,423 / 1,960 ± 644

4.79

0.00 ± 0.00 / 0.00 ± 0.00

0.00 ± 0.00 / 17.05 ± 0.35

0.00 ± 0.00 / 33.34 ± 0.31

0.00 ± 0.00 / 0.02 ± 0.01

37.66 ± 0.76 / 0.15 ± 0.02

0.87 ± 0.51 / 24.67 ± 0.61

1.95 ± 1.33 / 25.62 ± 0.89

13.0.0

ai-forever/mGPT (few-shot)

unknown

100

2077

True

11,734 ± 3,124 / 2,174 ± 720

4.96

0.00 ± 0.00 / 0.00 ± 0.00

0.19 ± 1.24 / 17.20 ± 1.22

-0.12 ± 0.91 / 36.65 ± 3.92

0.00 ± 0.00 / 1.29 ± 0.49

29.43 ± 1.80 / 2.35 ± 0.59

-0.69 ± 0.79 / 22.79 ± 0.51

0.15 ± 0.35 / 24.25 ± 0.59

12.10.0