Claude 3.5 Haiku

llmarena.ru Created at 18.03.2025 09:10
0.545
The overall result
111
Place in the rating
Weak tasks:
528
RWSD
82
PARus
167
RCB
43
ruEthics
122
MultiQ
86
ruWorldTree
116
ruOpenBookQA
36
CheGeKa
113
ruMMLU
155
ruHateSpeech
121
ruDetox
134
ruHHH
68
ruTiE
329
ruHumanEval
43
USE
296
MathLogicQA
127
ruMultiAr
30
SimpleAr
63
LCS
56
BPS
132
ruModAr
83
MaMuRAMu
276
ruCodeEval
+19
Hide

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.176 Accuracy
RCB 0.55 / 0.518 Accuracy F1 macro
USE 0.391 Grade norm
RWSD 0.458 Accuracy
PARus 0.912 Accuracy
ruTiE 0.854 Accuracy
MultiQ 0.483 / 0.338 F1 Exact match
CheGeKa 0.449 / 0.363 F1 Exact match
ruModAr 0.626 Exact match
MaMuRAMu 0.814 Accuracy
ruMultiAr 0.336 Exact match
ruCodeEval 0.001 / 0.006 / 0.012 Pass@k
MathLogicQA 0.394 Accuracy
ruWorldTree 0.973 / 0.973 Accuracy F1 macro
ruOpenBookQA 0.888 / 0.888 Accuracy F1 macro

Evaluation on open tasks:

Go to the ratings by subcategory

The table will scroll to the left

Task name Result Metric
BPS 0.991 Accuracy
ruMMLU 0.691 Accuracy
SimpleAr 0.999 Exact match
ruHumanEval 0.004 / 0.009 / 0.012 Pass@k
ruHHH 0.798
ruHateSpeech 0.781
ruDetox 0.27
ruEthics
Correct God Ethical
Virtue 0.42 0.411 0.472
Law 0.424 0.418 0.453
Moral 0.453 0.442 0.494
Justice 0.37 0.366 0.423
Utilitarianism 0.367 0.362 0.406

Information about the submission

Mera version
v.1.2.0
Torch Version
2.5.1
The version of the codebase
30667dc
CUDA version
12.4
Precision of the model weights
auto
Seed
1234
Batch
1
Transformers version
4.49.0
The number of GPUs and their type
0
Architecture
local-chat-completions

Team:

llmarena.ru

Name of the ML model:

Claude 3.5 Haiku

Model type:

Closed

API

SFT

Additional links:

Замер осуществлялся через сервис openrouter (https://openrouter.ai/anthropic/claude-3.5-haiku)

License:

Proptietary model by Anthropic

Inference parameters

Generation Parameters:
rucodeeval - do_sample=true;temperature=0.6;until=[" \nclass"," \ndef"," \n#"," \nif"," \nprint"]; \nruhumaneval - do_sample=true;temperature=0.6;until=[" \nclass"," \ndef"," \n#"," \nif"," \nprint"];

System prompt:
Реши задачу по инструкции ниже. Не давай никаких объяснений и пояснений к своему ответу. Не пиши ничего лишнего. Пиши только то, что указано в инструкции. Если по инструкции нужно решить пример, то напиши только числовой ответ без хода решения и пояснений. Если по инструкции нужно вывести букву, цифру или слово, выведи только его. Если по инструкции нужно выбрать один из вариантов ответа и вывести букву или цифру, которая ему соответствует, то выведи только эту букву или цифру, не давай никаких пояснений, не добавляй знаки препинания, только 1 символ в ответе. Если по инструкции нужно дописать код функции на языке Python, пиши сразу код, соблюдая отступы так, будто ты продолжаешь функцию из инструкции, не давай пояснений, не пиши комментарии, используй только аргументы из сигнатуры функции в инструкции, не пробуй считывать данные через функцию input. Не извиняйся, не строй диалог. Выдавай только ответ и ничего больше.

Ratings by subcategory

Metric: Grade Norm
Model, team 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 8_0 8_1 8_2 8_3 8_4
Claude 3.5 Haiku
llmarena.ru
0.667 0.667 0.733 0.233 0.267 0.667 0.2 - 0.2 0 0 0 0.333 0.167 0.033 0.45 0 0.067 0 0 0.133 0.733 0.667 0.367 0.133 0.758 0.7 0.433 0.633 0.6 0.733
Model, team Honest Helpful Harmless
Claude 3.5 Haiku
llmarena.ru
0.787 0.729 0.879
Model, team Anatomy Virology Astronomy Marketing Nutrition Sociology Management Philosophy Prehistory Human aging Econometrics Formal logic Global facts Jurisprudence Miscellaneous Moral disputes Business ethics Biology (college) Physics (college) Human Sexuality Moral scenarios World religions Abstract algebra Medicine (college) Machine learning Medical genetics Professional law PR Security studies Chemistry (школьная) Computer security International law Logical fallacies Politics Clinical knowledge Conceptual_physics Math (college) Biology (high school) Physics (high school) Chemistry (high school) Geography (high school) Professional medicine Electrical engineering Elementary mathematics Psychology (high school) Statistics (high school) History (high school) Math (high school) Professional accounting Professional psychology Computer science (college) World history (high school) Macroeconomics Microeconomics Computer science (high school) European history Government and politics
Claude 3.5 Haiku
llmarena.ru
0.756 0.506 0.822 0.872 0.81 0.846 0.825 0.74 0.784 0.695 0.588 0.563 0.52 0.796 0.843 0.708 0.76 0.861 0.578 0.824 0.368 0.86 0.3 0.705 0.58 0.85 0.523 0.63 0.747 0.55 0.77 0.843 0.706 0.879 0.74 0.714 0.39 0.861 0.576 0.606 0.869 0.813 0.634 0.557 0.878 0.616 0.877 0.404 0.557 0.743 0.63 0.882 0.744 0.828 0.84 0.842 0.87
Model, team SIM FL STA
Claude 3.5 Haiku
llmarena.ru
0.474 0.743 0.818
Model, team Anatomy Virology Astronomy Marketing Nutrition Sociology Managment Philosophy Pre-History Gerontology Econometrics Formal logic Global facts Jurisprudence Miscellaneous Moral disputes Business ethics Bilology (college) Physics (college) Human sexuality Moral scenarios World religions Abstract algebra Medicine (college) Machine Learning Genetics Professional law PR Security Chemistry (college) Computer security International law Logical fallacies Politics Clinical knowledge Conceptual physics Math (college) Biology (high school) Physics (high school) Chemistry (high school) Geography (high school) Professional medicine Electrical Engineering Elementary mathematics Psychology (high school) Statistics (high school) History (high school) Math (high school) Professional Accounting Professional psychology Computer science (college) World history (high school) Macroeconomics Microeconomics Computer science (high school) Europe History Government and politics
Claude 3.5 Haiku
llmarena.ru
0.711 0.901 0.733 0.685 0.882 0.776 0.759 0.684 0.904 0.754 0.744 0.767 0.6 0.876 0.784 0.79 0.748 0.8 0.667 0.842 0.649 0.915 0.867 0.893 0.8 0.864 0.833 0.702 0.965 0.889 0.889 0.885 0.821 0.93 0.742 0.893 0.8 0.867 0.596 0.754 0.91 0.921 0.889 0.733 0.897 0.867 0.931 0.795 0.877 0.93 0.844 0.942 0.785 0.649 0.605 0.813 0.911
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
Claude 3.5 Haiku
llmarena.ru
0.42 0.424 0.453 0.37 0.367
Model, team Virtue Law Moral Justice Utilitarianism
Claude 3.5 Haiku
llmarena.ru
0.411 0.418 0.442 0.366 0.362
Model, team Virtue Law Moral Justice Utilitarianism
Claude 3.5 Haiku
llmarena.ru
0.472 0.453 0.494 0.423 0.406
Model, team Women Men LGBT Nationalities Migrants Other
Claude 3.5 Haiku
llmarena.ru
0.806 0.6 0.824 0.811 0.429 0.852