Olmo-3.1-32B-Instruct

MERA Created at 18.03.2026 08:23
0.434
The overall result
276
Place in the rating
Weak tasks:
338
RWSD
281
PARus
243
RCB
205
ruEthics
183
MultiQ
276
ruWorldTree
236
ruOpenBookQA
296
CheGeKa
285
ruMMLU
329
ruHateSpeech
164
ruDetox
226
ruHHH
172
ruTiE
249
ruHumanEval
197
USE
119
MathLogicQA
586
ruMultiAr
53
LCS
356
BPS
210
ruModAr
347
MaMuRAMu
262
ruCodeEval
+18
Hide

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.188 Accuracy
RCB 0.534 / 0.51 Accuracy F1 macro
USE 0.182 Grade norm
RWSD 0.515 Accuracy
PARus 0.774 Accuracy
ruTiE 0.736 Accuracy
MultiQ 0.439 / 0.293 F1 Exact match
CheGeKa 0.094 / 0.06 F1 Exact match
ruModAr 0.503 Exact match
MaMuRAMu 0.561 Accuracy
ruMultiAr 0.003 Exact match
ruCodeEval 0.002 / 0.008 / 0.012 Pass@k
MathLogicQA 0.517 Accuracy
ruWorldTree 0.848 / 0.679 Accuracy F1 macro
ruOpenBookQA 0.798 / 0.798 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 0.857 Accuracy
ruMMLU 0.527 Accuracy
SimpleAr 0.0 Exact match
ruHumanEval 0.013 / 0.026 / 0.037 Pass@k
ruHHH 0.708
ruHateSpeech 0.645
ruDetox 0.228
ruEthics
Correct God Ethical
Virtue 0.32 0.35 0.451
Law 0.304 0.325 0.415
Moral 0.333 0.358 0.47
Justice 0.266 0.287 0.399
Utilitarianism 0.272 0.286 0.386

Information about the submission

Mera version
v1.2.0
Torch Version
2.10.0
The version of the codebase
0ac3a14
CUDA version
12.8
Precision of the model weights
bfloat16
Seed
1234
Batch
1
Transformers version
4.57.6
The number of GPUs and their type
4 x NVIDIA A100-SXM4-80GB
Architecture
vllm

Team:

MERA

Name of the ML model:

Olmo-3.1-32B-Instruct

Model size

32.0B

Model type:

Opened

SFT

Additional links:

https://arxiv.org/abs/2512.13961

Architecture description:

Model Description Developed by: Allen Institute for AI (Ai2) Model type: a Transformer style autoregressive language model. Language(s) (NLP): English

Description of the training:

Model Details Stage 1: SFT supervised fine-tuning on the Dolci-Think-SFT-7B dataset. This dataset consits of math, code, chat, and general knowledge queries. Datasets: Dolci-Think-SFT-7B, Dolci-Instruct-SFT Stage 2:DPO direct preference optimization on the Dolci-Think-DPO-7B dataset. This dataset consits of math, code, chat, and general knowledge queries. Datasets: Dolci-Think-DPO-7B, Dolci-Instruct-DPO Stage 3: RLVR reinforcement learning from verifiable rewards on the Dolci-Think-RL-7B dataset. This dataset consits of math, code, instruction-following, and general chat queries. Datasets: Dolci-Think-RL-7B, Dolci-Instruct-RL

Pretrain data:

Dec. 2024

License:

Apache 2.0

Inference parameters

Generation Parameters:
simplear - do_sample=false;until=["\n"]; \nchegeka - do_sample=false;until=["\n"]; \nrudetox - do_sample=false;until=["\n"]; \nrumultiar - do_sample=false;until=["\n"]; \nuse - do_sample=false;until=["\n","."]; \nmultiq - do_sample=false;until=["\n"]; \nrumodar - do_sample=false;until=["\n"]; \nruhumaneval - do_sample=true;temperature=0.6;until=["\nclass","\ndef","\n#","\nif","\nprint"]; \nrucodeeval - do_sample=true;temperature=0.6;until=["\nclass","\ndef","\n#","\nif","\nprint"];

The size of the context:
simplear, chegeka, rudetox, rumultiar, use, multiq, rumodar, ruhumaneval, rucodeeval - 65536

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:
{%- set has_system = messages|selectattr('role', 'equalto', 'system')|list|length > 0 -%}{%- if not has_system -%}{{- '<|im_start|>system \nYou are Olmo, a helpful AI assistant built by Ai2. Your date cutoff is December 2024, and your model weights are available at https://huggingface.co/allenai. ' -}}{%- if tools is none or (tools | length) == 0 -%}{{- 'You do not currently have access to any functions. <functions></functions><|im_end|> \n' -}}{%- else -%}{{- 'You are provided with function signatures within <functions></functions> XML tags. You may call one or more functions to assist with the user query. Output any function calls within <function_calls></function_calls> XML tags. Do not make assumptions about what values to plug into functions.' -}}{{- '<functions>' -}}{{- tools | tojson -}}{{- '</functions><|im_end|> \n' -}}{%- endif -%}{%- endif -%}{%- for message in messages -%}{%- if message['role'] == 'system' -%}{{- '<|im_start|>system \n' + message['content'] -}}{%- if tools is not none -%}{{- '<functions>' -}}{{- tools | tojson -}}{{- '</functions>' -}}{%- elif message.get('functions', none) is not none -%}{{- ' <functions>' + message['functions'] + '</functions>' -}}{%- endif -%}{{- '<|im_end|> \n' -}}{%- elif message['role'] == 'user' -%}{{- '<|im_start|>user \n' + message['content'] + '<|im_end|> \n' -}}{%- elif message['role'] == 'assistant' -%}{{- '<|im_start|>assistant \n' -}}{%- if message.get('content', none) is not none -%}{{- message['content'] -}}{%- endif -%}{%- if message.get('function_calls', none) is not none -%}{{- '<function_calls>' + message['function_calls'] + '</function_calls>' -}}{% elif message.get('tool_calls', none) is not none %}{{- '<function_calls>' -}}{%- for tool_call in message['tool_calls'] %}{%- if tool_call is mapping and tool_call.get('function', none) is not none %}{%- set args = tool_call['function']['arguments'] -%}{%- set ns = namespace(arguments_list=[]) -%}{%- for key, value in args.items() -%}{%- set ns.arguments_list = ns.arguments_list + [key ~ '=' ~ (value | tojson)] -%}{%- endfor -%}{%- set arguments = ns.arguments_list | join(', ') -%}{{- tool_call['function']['name'] + '(' + arguments + ')' -}}{%- if not loop.last -%}{{ ' \n' }}{%- endif -%}{% else %}{{- tool_call -}}{%- endif %}{%- endfor %}{{- '</function_calls>' -}}{%- endif -%}{%- if not loop.last -%}{{- '<|im_end|>' + ' \n' -}}{%- else -%}{{- eos_token -}}{%- endif -%}{%- elif message['role'] == 'environment' -%}{{- '<|im_start|>environment \n' + message['content'] + '<|im_end|> \n' -}}{%- elif message['role'] == 'tool' -%}{{- '<|im_start|>environment \n' + message['content'] + '<|im_end|> \n' -}}{%- endif -%}{%- if loop.last and add_generation_prompt -%}{{- '<|im_start|>assistant\n' -}}{%- endif -%}{%- endfor -%}

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
Olmo-3.1-32B-Instruct
MERA
0.1 0.2 0.633 0.233 0 0.133 0 - 0 0 0 0.033 0.2 0.1 0.033 0.317 0.033 0 0.033 0 0 0.4 0.233 0.1 0.1 0.458 0.133 0.167 0.233 0.333 0.3
Model, team Honest Helpful Harmless
Olmo-3.1-32B-Instruct
MERA
0.672 0.729 0.724
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
Olmo-3.1-32B-Instruct
MERA
0.407 0.488 0.539 0.722 0.598 0.677 0.65 0.559 0.509 0.596 0.404 0.516 0.42 0.611 0.534 0.604 0.63 0.438 0.511 0.534 0.372 0.491 0.31 0.462 0.527 0.5 0.398 0.565 0.624 0.37 0.62 0.744 0.589 0.667 0.509 0.603 0.45 0.632 0.503 0.542 0.626 0.386 0.51 0.605 0.646 0.528 0.578 0.481 0.426 0.5 0.55 0.7 0.544 0.576 0.85 0.63 0.596
Model, team SIM FL STA
Olmo-3.1-32B-Instruct
MERA
0.476 0.653 0.787
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
Olmo-3.1-32B-Instruct
MERA
0.356 0.535 0.483 0.556 0.5 0.603 0.5 0.474 0.346 0.477 0.615 0.617 0.342 0.605 0.404 0.457 0.645 0.467 0.456 0.509 0.509 0.678 0.667 0.568 0.778 0.47 0.654 0.386 0.825 0.6 0.822 0.615 0.589 0.649 0.5 0.607 0.6 0.644 0.439 0.554 0.584 0.365 0.8 0.778 0.828 0.844 0.569 0.773 0.631 0.737 0.889 0.333 0.696 0.61 0.651 0.327 0.6
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
Olmo-3.1-32B-Instruct
MERA
0.32 0.304 0.333 0.266 0.272
Model, team Virtue Law Moral Justice Utilitarianism
Olmo-3.1-32B-Instruct
MERA
0.35 0.325 0.358 0.287 0.286
Model, team Virtue Law Moral Justice Utilitarianism
Olmo-3.1-32B-Instruct
MERA
0.451 0.415 0.47 0.399 0.386
Model, team Women Men LGBT Nationalities Migrants Other
Olmo-3.1-32B-Instruct
MERA
0.722 0.571 0.529 0.541 0.429 0.672