Papers › Measuring Massive Multitask Language Understanding

Measuring Massive Multitask Language Understanding

7 Sep 2020arXiv:2009.03300archive 2025-07-28

Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, Jacob Steinhardt

We propose a new test to measure a text model's multitask accuracy. The test covers 57 tasks including elementary mathematics, US history, computer science, law, and more. To attain high accuracy on this test, models must possess extensive world knowledge and problem solving ability. We find that while most recent models have near random-chance accuracy, the very largest GPT-3 model improves over random chance by almost 20 percentage points on average. However, on every one of the 57 tasks, the best models still need substantial improvements before they can reach expert-level accuracy. Models also have lopsided performance and frequently do not know when they are wrong. Worse, they still have near-random accuracy on some socially important subjects such as morality and law. By comprehensively evaluating the breadth and depth of a model's academic and professional understanding, our test can be used to analyze models across many tasks and to identify important shortcomings.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

No sample from the archive's official repository (hendrycks/test) is linked to this paper; the 26 samples below come from 3 community repositories. Syntology harvested code from hendrycks/test for another paper.

hendrycks/test: Syntology's graph links these samples to MRAG: Benchmarking Retrieval-Augmented Generation for Bio-medicine (arXiv:2601.16503; that paper's own run record: 9 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 8 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method.

Syntology 26 code samples harvested from 3 repositories linked to this paper: 19 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 9 where Syntology's instrument failed) · 7 unverified. Where Syntology's instrument failed: 8 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 26 samples from 3 repositories, 19 ran (10 with no instrument failure; 9 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 10 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 3 of the 19 samples that ran have no recorded date.

18 repositories listed; official and paper-mentioned ones first.

hendrycks/test officialmentioned in papermentioned on GitHubSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to MRAG: Benchmarking Retrieval-Augmented Generation for Bio-medicine (arXiv:2601.16503; that paper's own run record: 9 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 8 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
8023looker/med-rr mentioned on GitHubpytorchSyntology: not harvested report
aryopg/mmlu-redux mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (1 instrument-failed) · 0 unverified report
baichuan-inc/baichuan-13b mentioned on GitHubpytorchApache-2.0Syntology: not harvested report
baichuan-inc/baichuan-7b mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to RoFormer: Enhanced Transformer with Rotary Position Embedding (arXiv:2104.09864; that paper's own run record: 14 ran (of which 5 constructed an object rather than computing a result; 8 with no instrument failure: 1 honoured, 0 violated, 7 with no contract checked; 6 where Syntology's instrument failed) · 2 unverified); nothing checks that they implement this paper's method. report
baichuan-inc/baichuan2 mentioned on GitHubpytorchApache-2.0Syntology: not harvested report
bunsenfeng/knowledge_card mentioned on GitHubpytorchMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Knowledge Card: Filling LLMs' Knowledge Gaps with Plug-in Specialized Language Models (arXiv:2305.09955; that paper's own run record: 3 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 2 where Syntology's instrument failed) · 5 unverified); nothing checks that they implement this paper's method. report
duxiaoman-di/xuanyuan mentioned on GitHubpytorchSyntology: not harvested report
instructlab/instructlab mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to LAB: Large-Scale Alignment for ChatBots (arXiv:2403.01081; that paper's own run record: 4 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
lm-sys/routellm mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to RouteLLM: Learning to Route LLMs with Preference Data (arXiv:2406.18665; that paper's own run record: 5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 2 unverified); nothing checks that they implement this paper's method. report
nlp-uoregon/mlmm-evaluation mentioned on GitHubApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Okapi: Instruction-tuned Large Language Models in Multiple Languages with... (arXiv:2307.16039; that paper's own run record: 8 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 1 violated, 0 with no contract checked; 6 where Syntology's instrument failed) · 3 unverified); nothing checks that they implement this paper's method. report
ollmer/mmlu mentioned on GitHubMITSyntology: 9 samples, 9 ran (8 instrument-failed) · 0 unverified report
opennlplab/transnormerllm mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Various Lengths, Constant Speed: Efficient Language Modeling with Lightning Attention (arXiv:2405.17381; that paper's own run record: 5 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
petezh/ChatGPT-Advice mentioned on GitHubMITSyntology: not harvested report
sanjass/MultiTaskLearning mentioned on GitHubpytorchMITSyntology: 16 samples, 9 ran (0 instrument-failed) · 7 unverified report
vivo-ai-lab/bluelm mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to LoRA: Low-Rank Adaptation of Large Language Models (arXiv:2106.09685; that paper's own run record: 51 ran (of which 19 constructed an object rather than computing a result; 43 with no instrument failure: 1 honoured, 0 violated, 42 with no contract checked; 8 where Syntology's instrument failed) · 33 unverified); nothing checks that they implement this paper's method. report
xverse-ai/xverse-13b mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Llama 2: Open Foundation and Fine-Tuned Chat Models (arXiv:2307.09288; that paper's own run record: 33 ran (of which 7 constructed an object rather than computing a result; 20 with no instrument failure: 1 honoured, 1 violated, 18 with no contract checked; 13 where Syntology's instrument failed) · 19 unverified); nothing checks that they implement this paper's method. report
xverse-ai/xverse-7b mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to ACL Anthology id 2024.findings-acl.240 (that paper's own run record: 0 ran · 2 unverified); nothing checks that they implement this paper's method. report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

26 samples harvested: 19 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 9 where Syntology's instrument failed) · 7 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
8ran · our draft was wrong
1ran; the contract check could not run
9ran
7unverified

Licence: 1 of the 26 samples is pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 3 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

aryopg/mmlu-redux community (archive-listed) 1 sample, 1 ran (1 instrument-failed) · 0 unverified
create_model_path sft/sft_full.py ran · our draft was wrong CC-BY-4.0 · pointer only · 7d2a2984ad40114e · report fixture: signature args:Args/0/? · returned str · level recorded 2026-09-22
ollmer/mmlu community (archive-listed) 9 samples, 9 ran (8 instrument-failed) · 0 unverified
softmax in 240 papers
census: 53 implementations from 63 papers, 16 distinct outputs across 8 buckets, one shared input per bucket this sample: bucket 2 of 8, arg 1: rank 1, kind float, dtype float64
evaluate.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted MIT (permissive) · 19f01570b0a18e2b · report fixture: signature x:ndarray/1/float · returned ndarray · level recorded, date not recorded
format_example in 43 papers evaluate.py ran · our draft was wrong MIT (permissive) · cd763eaf1ac287e7 · report fixture: signature df:DataFrame/0/?, idx:int/0/int, include_answer:bool/0/bool · returned str · level recorded, date not recorded
format_subject in 26 papers evaluate.py ran · our draft was wrong fingerprinted MIT (permissive) · 6ab745408cb8648b · report fixture: signature subject:str/0/str · returned str · level recorded, date not recorded
gen_prompt in 35 papers evaluate_flan.py ran · our draft was wrong MIT (permissive) · 3383488ea82b6c15 · report fixture: signature train_df:DataFrame/0/?, subject:str/0/str, k:int/0/int · returned str · level recorded 2026-09-24
get_adaptive_bins calib_tools.py ran · our draft was wrong fingerprinted MIT (permissive) · 432838d3c5f2ab75 · report fixture: signature predictions:ndarray/1/float, num_bins:int/0/int · returned ndarray · level recorded 2026-09-26
get_encoder in 45 papers crop.py ran · our draft was wrong MIT (permissive) · 3c57d0ee2854879f · report fixture: signature model_name:str/0/str · returned Encoder · level recorded 2026-09-26
get_pairs in 260 papers crop.py ran · our draft was wrong MIT (permissive) · 4f76173dfe4b0c11 · report fixture: signature word:tuple/1/str · returned set · level recorded 2026-09-26
mean in 84 papers calib_tools.py ran · honoured contract fingerprinted MIT (permissive) · 0af29c87aa4bd6a8 · report fixture: signature inputs:ndarray/1/float · returned float64 · level recorded 2026-09-26
one_hot_encode in 18 papers calib_tools.py ran · our draft was wrong MIT (permissive) · 77933604fbca56e7 · report fixture: signature labels:ndarray/1/int, num_classes:int/0/int · returned ndarray · level recorded 2026-09-26
sanjass/MultiTaskLearning community (archive-listed) 16 samples, 9 ran (0 instrument-failed) · 7 unverified
check_contains_num ex2_data_parsing/data_parse.py ran MIT (permissive) · 3a55bded83f59394 · report fixture: signature question:list/1/str · returned bool · level recorded 2026-09-24
check_numeric ex2_data_parsing/data_parse.py ran MIT (permissive) · e52cf9ef3b07a19c · report fixture: signature question:list/1/str · returned bool · level recorded 2026-09-24
format_exp 036questions/utils.py ran fingerprinted MIT (permissive) · 0a261ea0770f9244 · report fixture: signature x:int/0/int · returned str · level recorded 2026-09-24
format_num 036questions/utils.py ran fingerprinted MIT (permissive) · 4839a2077dc13fca · report fixture: signature x:int/0/int · returned str · level recorded 2026-09-24
gen_quant_cells ex2_data_parsing/data_parse.py ran MIT (permissive) · 29d8c9d894d5cb2e · report fixture: signature question:list/1/str · returned list · level recorded 2026-09-24
get_and_update 036questions/writedata.py ran MIT (permissive) · 0ae544dd79bf6e53 · report fixture: signature mod:callable/0/?, train_id:int/0/int, test_id:int/0/int, train_data:list/1/empty, test_data:list/1/empty, test_answers:list/1/empty · returned tuple · level recorded 2026-09-24
get_quant_cells 036questions/utils.py ran fingerprinted MIT (permissive) · 0e51b6d0af943f39 · report fixture: signature question:str/0/str · returned list · level recorded 2026-09-24
infix_to_prefix exercise2/util.py ran fingerprinted MIT (permissive) · 87d40f0c0a1ce3ac · report fixture: signature expression:str/0/str · returned list · level recorded 2026-09-24
tokenize_and_separate_quants exercise2/util.py ran MIT (permissive) · 4e014ec57f85bafe · report fixture: signature data:list/1/dict, n_min_vocab:int/0/int · returned tuple · level recorded 2026-09-24
return_data in 6 papers 036questions/filtering.py unverified MIT (permissive) · 69e44b454bef1c76 · report
return_data in 6 papers 036questions/filtering2.py unverified MIT (permissive) · b441fdbef0a06567 · report
return_data in 6 papers 036questions/gradientdescent.py unverified MIT (permissive) · bddd1344d4501303 · report
return_data in 6 papers 036questions/gradientdescent2.py unverified MIT (permissive) · 5dfa113ed2d4f658 · report
return_data in 6 papers 036questions/logisticregression.py unverified MIT (permissive) · de6faacbeb0eb06d · report
return_data in 6 papers 036questions/rnn.py unverified MIT (permissive) · 69930c08eceb6a88 · report
setup in 24 papers exercise2/util.py unverified MIT (permissive) · 26f49fbc83128412 · report

Tasks

Elementary MathematicsMulti-Task LearningMulti-task Language UnderstandingWorld Knowledge

Datasets

Introduced by this paper, per the archive.

MMLMMLU

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

AdamAttentionAttention DropoutBPECosine AnnealingDense ConnectionsDropoutGPT-3Layer NormalizationLinear LayerLinear Warmup With Cosine AnnealingMulti-Head AttentionResidual ConnectionSoftmaxWeight Decay

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections