Papers › The Curious Case of Neural Text Degeneration

The Curious Case of Neural Text Degeneration

22 Apr 2019ICLR 2020 1arXiv:1904.09751archive 2025-07-28

Ari Holtzman, Jan Buys, Li Du, Maxwell Forbes, Yejin Choi

Despite considerable advancements with deep neural language models, the enigma of neural text degeneration persists when these models are tested as text generators. The counter-intuitive empirical observation is that even though the use of likelihood as training objective leads to high quality models for a broad range of language understanding tasks, using likelihood as a decoding objective leads to text that is bland and strangely repetitive. In this paper, we reveal surprising distributional differences between human text and machine text. In addition, we find that decoding strategies alone can dramatically effect the quality of machine text, even when generated from exactly the same neural language model. Our findings motivate Nucleus Sampling, a simple but effective method to draw the best out of neural generation. By sampling text from the dynamic nucleus of the probability distribution, which allows for diversity while effectively truncating the less reliable tail of the distribution, the resulting text better demonstrates the quality of human text, yielding enhanced diversity without sacrificing fluency and coherence.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

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

Syntology 17 code samples harvested from 7 repositories linked to this paper: 10 ran (of which 1 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 0 violated, 4 with no contract checked; 5 where Syntology's instrument failed) · 7 unverified. Where Syntology's instrument failed: 4 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: official repository: 6 samples from 1 repository, 4 ran (0 with no instrument failure; 4 where Syntology's instrument failed); community (archive-listed): 11 samples from 6 repositories, 6 ran (5 with no instrument failure; 1 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 4 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

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

ari-holtzman/degen officialmentioned in paperpytorchSyntology: 6 samples, 4 ran (4 instrument-failed) · 2 unverified report
Advancing-Machine-Human-Reasoning-Lab/apt mentioned on GitHubpytorchSyntology: not harvested report
AleksSol/mmp_activity_diary mentioned on GitHubSyntology: not harvested report
THUDM/KOBE mentioned on GitHubpytorchMITSyntology: 2 samples, 2 ran (0 instrument-failed) · 0 unverified report
THUcqb/KOBE mentioned on GitHubpytorchMITSyntology: not harvested report
TaiseiAso/BiGruAttEncDec mentioned on GitHubpytorchSyntology: 4 samples, 0 ran · 4 unverified report
c00k1ez/plain-transformers mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Neural Text Generation with Unlikelihood Training (arXiv:1908.04319; that paper's own run record: 6 ran (of which 1 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 4 where Syntology's instrument failed) · 0 unverified) and Understanding the Difficulty of Training Transformers (arXiv:2004.08249; that paper's own run record: 4 ran (of which 2 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 1 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
colorfulscoop/tfdlg mentioned on GitHubtfMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to On Layer Normalization in the Transformer Architecture (arXiv:2002.04745; that paper's own run record: 10 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 2 honoured, 0 violated, 8 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
microsoft/kid mentioned on GitHubpytorchMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Knowledge Infused Decoding (arXiv:2204.03084; that paper's own run record: 3 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
noriyukipy/tfchat mentioned on GitHubtfMITSyntology: not harvested report
noriyukipy/tfdlg mentioned on GitHubtfSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
prakhar21/textaugmentation-gpt2 mentioned on GitHubpytorchMITSyntology: not harvested report
qibinc/KOBE mentioned on GitHubpytorchMITSyntology: not harvested report
sander102907/autoencoder_program_synthesis mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
thu-coai/CDial-GPT mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (1 instrument-failed) · 0 unverified report
MindSpore-scientific/code-8 mindsporeSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to SimBa: Simplicity Bias for Scaling Up Parameters in Deep Reinforcement Learning (arXiv:2410.09754; that paper's own run record: 6 ran (of which 2 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
labmlai/annotated_deep_learning_paper_implementations pytorchSyntology: 2 samples, 1 ran (0 instrument-failed) · 1 unverified 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

17 samples harvested: 10 ran (of which 1 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 0 violated, 4 with no contract checked; 5 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 has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
4ran · our draft was wrong
1ran; the contract check could not run
4ran
7unverified

Licence: 11 of the 17 samples are 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 7 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. 1 sample whose link to this paper records no repository is listed under one of this paper's own repositories, because identical code was harvested from it for another 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”. For 1 sample that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “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.

ari-holtzman/degen official repository 6 samples, 4 ran (4 instrument-failed) · 2 unverified
_gumbel gen.py ran · our draft was wrong fingerprinted GPL-3.0 (copyleft) · pointer only · 2d61025ad472d12d · report fixture: signature u:Tensor/2/float · returned Tensor · level recorded 2026-09-22
_shift_gumbel_maximum gen.py ran · our draft was wrong fingerprinted GPL-3.0 (copyleft) · pointer only · 7c5ca0a2e17c5e6b · report fixture: signature g_phi:Tensor/2/float, T:Tensor/1/float, dim:int/0/int, Z:NoneType/0/? · returned Tensor · level recorded 2026-09-22
gumbel_sbs_decode gen.py ran · our draft was wrong GPL-3.0 (copyleft) · pointer only · 46d2256fd27f9ce4 · report fixture: signature model:callable/0/?, init:tuple/1/Tensor, w:int/0/int, max_len:int/0/int, sep:int/0/int, device:str/0/str, batch_size:int/0/int · returned list · level recorded 2026-09-22
gumbel_with_maximum gen.py ran · our draft was wrong fingerprinted GPL-3.0 (copyleft) · pointer only · 975e63af902daf2e · report fixture: signature phi:Tensor/2/float, T:Tensor/1/float, dim:int/0/int · returned tuple · level recorded 2026-09-22
gumbel gen.py unverified GPL-3.0 (copyleft) · pointer only · c8316a45e99146db · report
gumbel_like gen.py unverified GPL-3.0 (copyleft) · pointer only · 6c601d82e95610de · report
THUDM/KOBE community (archive-listed) 2 samples, 2 ran (0 instrument-failed) · 0 unverified
accuracy in 872 papers
census: 8 implementations from 7 papers, 7 distinct outputs across 6 buckets, one shared input per bucket
kobe/utils/helpers.py ran MIT (permissive) · 6135894d997967b3 · report fixture: signature logits:Tensor/2/float, targets:Tensor/1/int · returned float · level recorded 2026-09-24
get_bert_vocab_size kobe/utils/helpers.py ran MIT (permissive) · b9ed4ea16c8288f0 · report fixture: signature vocab_path:str/0/str · returned int · level recorded 2026-09-24
TaiseiAso/BiGruAttEncDec community (archive-listed) 4 samples, 0 ran · 4 unverified
entity_enhance decode.py unverified no licence file found · pointer only · 717c7692322af957 · report
inf_suppression decode.py unverified no licence file found · pointer only · 7b0560d1ecad407a · report
repetitive_suppression decode.py unverified no licence file found · pointer only · 4aa0814a647dbccb · report
top_p_sampling_search decode.py unverified no licence file found · pointer only · 2cd2fb7377190229 · report
noriyukipy/tfdlg community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
filter_to_topp tfdlg/generations.py ran · honoured contract fingerprinted MIT (permissive) · f10aa6a3b19fd64e · report fixture: signature top_p:float/0/float, dist:ndarray/2/float · returned ndarray · level recorded 2026-09-22
sander102907/autoencoder_program_synthesis community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
Sampling autoencoder_program_synthesis/utils/Sampling.py ran Apache-2.0 (permissive) · 27aeb9112fe4cf52 · report fixture: signature logits:Tensor/2/float, temperature:float/0/float, top_k:int/0/int, top_p:float/0/float · returned Tensor · level recorded 2026-09-24
thu-coai/CDial-GPT community (archive-listed) 1 sample, 1 ran (1 instrument-failed) · 0 unverified
top_filtering in 13 papers
census: 5 implementations from 9 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 1, kind float, dtype float32
infer.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own ran; the contract check could not run (RAISES) · metamorphic tier: invariant licence of this copy not recorded · 695eaa24be6f2a1d · report fixture: signature logits:Tensor/1/float, top_k:int/0/int, top_p:float/0/float, threshold:float/0/float, filter_value:float/0/float · returned Tensor · level recorded 2026-09-22
labmlai/annotated_deep_learning_paper_implementations community (archive-listed) 2 samples, 1 ran (0 instrument-failed) · 1 unverified
NucleusSampler labml_nn/sampling/nucleus.py ran fingerprinted MIT (permissive) · 5a0768b2c98ccd3a · report fixture: returned NucleusSampler · level recorded 2026-09-22
Sampler in 21 papers labml_nn/sampling/nucleus.py unverified MIT (permissive) · 7d8fed5b92965553 · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it. Siblings are assemblies lifted from this paper in the same batch for the same method: alternative readings kept side by side; none replaces or supersedes another.

6 assemblies on this paper: 6 harvested, 0 generated, listed in 1 groups (siblings lifted in one batch form one group).

6 sibling assemblies for Nucleus Sampling (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.

harvested Assembly for Nucleus Sampling (graph method: Nucleus Sampling). 2 bricks; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 6 siblings

  1. NucleusSampler entry · level 1 · ran · 5a0768b2c98ccd3a
  2. Sampler sampler · level 0 · unverified · 7d8fed5b92965553

harvested Assembly for Nucleus Sampling (graph method: Nucleus Sampling). 1 brick; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

one of 6 siblings

  1. top_filtering entry · level 1 · ran; the contract check could not run (RAISES) · 695eaa24be6f2a1d

harvested Assembly for Nucleus Sampling (graph method: Nucleus Sampling). 1 brick; individually on a fixture: 1 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; 0 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

one of 6 siblings

  1. filter_to_topp entry · level 1 · ran · honoured contract · f10aa6a3b19fd64e

harvested Assembly for Nucleus Sampling (graph method: Nucleus Sampling). 1 brick; individually on a fixture: 1 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; 0 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

one of 6 siblings

  1. Sampling entry · level 1 · ran · 27aeb9112fe4cf52

harvested Assembly for Nucleus Sampling (graph method: Nucleus Sampling). 4 bricks; individually on a fixture: 0 ran · 4 unverified. The assembly has not been run end-to-end.

one of 6 siblings

  1. top_p_sampling_search entry · level 0 · unverified · 2cd2fb7377190229
  2. entity_enhance helper · level 0 · unverified · 717c7692322af957
  3. inf_suppression helper · level 0 · unverified · 7b0560d1ecad407a
  4. repetitive_suppression helper · level 0 · unverified · 4aa0814a647dbccb

harvested Assembly for Nucleus Sampling (graph method: Nucleus Sampling). 6 bricks; individually on a fixture: 4 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

one of 6 siblings

  1. gumbel_sbs_decode entry · level 1 · ran · our draft was wrong · 46d2256fd27f9ce4
  2. gumbel helper · level 0 · unverified · c8316a45e99146db
  3. gumbel_like helper · level 0 · unverified · 6c601d82e95610de
  4. gumbel_with_maximum sampler · level 1 · ran · our draft was wrong · 975e63af902daf2e
  5. _gumbel helper · level 1 · ran · our draft was wrong · 2d61025ad472d12d
  6. _shift_gumbel_maximum helper · level 1 · ran · our draft was wrong · 7c5ca0a2e17c5e6b

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="1904.09751") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

DiversityLanguage ModelingLanguage Modelling

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

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