Qwen3-4B-Instruct-2507

MERA Created at 11.02.2026 05:08
0.468
The overall result
199
Place in the rating
Weak tasks:
562
RWSD
196
PARus
250
RCB
178
ruEthics
159
MultiQ
154
ruWorldTree
169
ruOpenBookQA
323
CheGeKa
197
ruMMLU
224
ruHateSpeech
201
ruDetox
224
ruHHH
133
ruTiE
257
ruHumanEval
130
USE
102
MathLogicQA
124
ruMultiAr
63
SimpleAr
606
LCS
98
BPS
332
ruModAr
212
MaMuRAMu
240
ruCodeEval
+19
Hide

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.016 Accuracy
RCB 0.532 / 0.478 Accuracy F1 macro
USE 0.259 Grade norm
RWSD 0.431 Accuracy
PARus 0.852 Accuracy
ruTiE 0.778 Accuracy
MultiQ 0.456 / 0.279 F1 Exact match
CheGeKa 0.072 / 0.046 F1 Exact match
ruModAr 0.423 Exact match
MaMuRAMu 0.706 Accuracy
ruMultiAr 0.338 Exact match
ruCodeEval 0.006 / 0.006 / 0.006 Pass@k
MathLogicQA 0.569 Accuracy
ruWorldTree 0.945 / 0.945 Accuracy F1 macro
ruOpenBookQA 0.858 / 0.687 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 0.981 Accuracy
ruMMLU 0.621 Accuracy
SimpleAr 0.995 Exact match
ruHumanEval 0.012 / 0.012 / 0.012 Pass@k
ruHHH 0.708
ruHateSpeech 0.743
ruDetox 0.205
ruEthics
Correct God Ethical
Virtue 0.358 0.402 0.379
Law 0.374 0.381 0.362
Moral 0.388 0.425 0.389
Justice 0.323 0.337 0.321
Utilitarianism 0.285 0.351 0.321

Information about the submission

Mera version
v1.2.0
Torch Version
2.8
The version of the codebase
7030e16
CUDA version
12.8
Precision of the model weights
bfloat16
Seed
1234
Batch
1
Transformers version
4.56.1
The number of GPUs and their type
1 x NVIDIA A100 80GB
Architecture
local-chat-completions

Team:

MERA

Name of the ML model:

Qwen3-4B-Instruct-2507

Model size

4.0B

Model type:

Opened

SFT

License:

apache-2.0

Inference parameters

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.

Description of the template:
{%- if tools %} {{- '<|im_start|>system' }} {%- if messages[0].role == 'system' %} {{- messages[0].content + '' }} {%- endif %} {{- "# ToolsYou may call one or more functions to assist with the user query.You are provided with function signatures within <tools></tools> XML tags:<tools>" }} {%- for tool in tools %} {{- "" }} {{- tool | tojson }} {%- endfor %} {{- "</tools>For each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:<tool_call>{"name": <function-name>, "arguments": <args-json-object>}</tool_call><|im_end|>" }} {%- else %} {%- if messages[0].role == 'system' %} {{- '<|im_start|>system' + messages[0].content + '<|im_end|>' }} {%- endif %} {%- endif %} {%- for message in messages %} {%- if message.content is string %} {%- set content = message.content %} {%- else %} {%- set content = '' %} {%- endif %} {%- if (message.role == "user") or (message.role == "system" and not loop.first) %} {{- '<|im_start|>' + message.role + '' + content + '<|im_end|>' + '' }} {%- elif message.role == "assistant" %} {{- '<|im_start|>' + message.role + '' + content }} {%- if message.tool_calls %} {%- for tool_call in message.tool_calls %} {%- if (loop.first and content) or (not loop.first) %} {{- '' }} {%- endif %} {%- if tool_call.function %} {%- set tool_call = tool_call.function %} {%- endif %} {{- '<tool_call>{"name": "' }} {{- tool_call.name }} {{- '", "arguments": ' }} {%- if tool_call.arguments is string %} {{- tool_call.arguments }} {%- else %} {{- tool_call.arguments | tojson }} {%- endif %} {{- '}</tool_call>' }} {%- endfor %} {%- endif %} {{- '<|im_end|>' }} {%- elif message.role == "tool" %} {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} {{- '<|im_start|>user' }} {%- endif %} {{- '<tool_response>' }} {{- content }} {{- '</tool_response>' }} {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} {{- '<|im_end|>' }} {%- endif %} {%- endif %} {%- endfor %} {%- if add_generation_prompt %} {{- '<|im_start|>assistant' }} {%- endif %}

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
Qwen3-4B-Instruct-2507
MERA
0.433 0.367 0.733 0.233 0 0.233 0 - 0 0.033 0.067 0.033 0.2 0.067 0.033 0.45 0.067 0.033 0 0 0.1 0.633 0.267 0.167 0.267 0.467 0.233 0.167 0.6 0.467 0.6
Model, team Honest Helpful Harmless
Qwen3-4B-Instruct-2507
MERA
0.639 0.678 0.81
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
Qwen3-4B-Instruct-2507
MERA
0.622 0.5 0.757 0.799 0.657 0.711 0.757 0.656 0.583 0.619 0.579 0.563 0.39 0.741 0.681 0.656 0.68 0.757 0.5 0.641 0.327 0.708 0.44 0.671 0.545 0.68 0.428 0.574 0.706 0.49 0.72 0.769 0.626 0.747 0.668 0.778 0.49 0.855 0.589 0.695 0.773 0.61 0.683 0.666 0.811 0.63 0.74 0.485 0.465 0.608 0.56 0.764 0.703 0.803 0.82 0.752 0.694
Model, team SIM FL STA
Qwen3-4B-Instruct-2507
MERA
0.689 0.573 0.591
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
Qwen3-4B-Instruct-2507
MERA
0.533 0.832 0.6 0.565 0.684 0.707 0.672 0.667 0.538 0.692 0.756 0.717 0.475 0.667 0.585 0.617 0.673 0.756 0.719 0.754 0.649 0.763 0.822 0.746 0.733 0.727 0.718 0.667 0.842 0.844 0.844 0.769 0.679 0.684 0.682 0.839 0.889 0.778 0.632 0.785 0.767 0.794 0.778 0.8 0.897 0.822 0.793 0.932 0.738 0.86 0.8 0.594 0.785 0.701 0.558 0.485 0.733
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
Qwen3-4B-Instruct-2507
MERA
0.358 0.374 0.388 0.323 0.285
Model, team Virtue Law Moral Justice Utilitarianism
Qwen3-4B-Instruct-2507
MERA
0.402 0.381 0.425 0.337 0.351
Model, team Virtue Law Moral Justice Utilitarianism
Qwen3-4B-Instruct-2507
MERA
0.379 0.362 0.389 0.321 0.321
Model, team Women Men LGBT Nationalities Migrants Other
Qwen3-4B-Instruct-2507
MERA
0.75 0.686 0.529 0.784 0.571 0.82