Browse › Computer Vision › Sentiment Analysis › IMDb
IMDb Benchmark (Sentiment Analysis)
Sentiment Analysis is the task of classifying the polarity of a given text. For instance, a text-based tweet can be categorized into either "positive", "negative", or "neutral". Given the text and accompanying labels, a model can be trained to predict the correct sentiment.
Sentiment Analysis techniques can be categorized into machine learning approaches, lexicon-based approaches, and even hybrid methods. Some subcategories of research in sentiment analysis include: multimodal sentiment analysis, aspect-based sentiment analysis, fine-grained opinion analysis, language specific sentiment analysis.
More recently, deep learning techniques, such as RoBERTa and T5, are used to train high-performing sentiment classifiers that are evaluated using metrics like F1, recall, and precision. To evaluate sentiment analysis systems, benchmark datasets like SST, GLUE, and IMDB movie reviews are used.
Further readings:
- Sentiment Analysis Based on Deep Learning: A Comparative Study
The archive carries no text for this table; the description above is the archive's text for the task Sentiment Analysis. archive 2025-07-28
Over time archive 2025-07-28
The chart needs JavaScript; the table below carries every value.
Direction inferred from the metric name, not from the archive: Accuracy (higher is better). Points are placed at the row's paper date; 48 of 49 rows carry one.
Results archive 2025-07-28
Archive rows end at the archive snapshot, 2025-07-28: no result published after that date is in this table. Rank is the archive's row order at that snapshot; not re-ranked here. Metric values are the archive's strings. Column headers sort the table in your browser; each row keeps its archive rank.
| Paper | Code | Ran Syntology | Report | |||||
|---|---|---|---|---|---|---|---|---|
| 1 | RoBERTa-large with LlamBERT | 96.68 | – | Paper | Code | 2024 | linked, not harvested | report |
| 2 | RoBERTa-large | 96.54 | – | Paper | Code | 2024 | linked, not harvested | report |
| 3 | XLNet | 96.21 | ✓ | Paper | Code | 2019 | 10 of 24 ran · 14 unverified | report |
| 4 | Heinsen Routing + RoBERTa Large | 96.2 | – | Paper | Code | 2022 | linked, not harvested | report |
| 5 | RoBERTa-large 355M + Entailment as Few-shot Learner | 96.1 | – | Paper | Code | 2021 | 1 of 3 ran · 2 unverified | report |
| 6 | GraphStar | 96.0 | – | Paper | Code | 2019 | 1 of 6 ran · 5 unverified | report |
| 7 | DV-ngrams-cosine with NB sub-sampling + RoBERTa.base | 95.94 | – | Paper | Code | 2022 | linked, not harvested | report |
| 8 | DV-ngrams-cosine + RoBERTa.base | 95.92 | – | Paper | Code | 2022 | linked, not harvested | report |
| 9 | Roberta_Large ST + Cosine Similarity Loss | 95.9 | – | – | – | not matched | report | |
| 10 | BERT large finetune UDA | 95.8 | ✓ | Paper | Code | 2019 | 15 of 52 ran · 37 unverified | report |
| 11 | BERT_large+ITPT | 95.79 | – | Paper | Code | 2019 | 6 of 18 ran · 12 unverified | report |
| 12 | RoBERTa.base | 95.79 | ✓ | Paper | Code | 2022 | linked, not harvested | report |
| 13 | L MIXED | 95.68 | – | Paper | Code | 2020 | 2 of 3 ran · 1 unverified | report |
| 14 | BERT_base+ITPT | 95.63 | – | Paper | Code | 2019 | 6 of 18 ran · 12 unverified | report |
| 15 | BERT large | 95.49 | – | Paper | Code | 2019 | 15 of 52 ran · 37 unverified | report |
| 16 | ULMFiT | 95.4 | – | Paper | Code | 2018 | 2 of 5 ran · 3 unverified | report |
| 17 | Llama-2-70b-chat (0-shot) | 95.39 | ✓ | Paper | Code | 2024 | linked, not harvested | report |
| 18 | FLAN 137B (few-shot, k=2) | 95 | ✓ | Paper | Code | 2021 | 0 of 1 ran · 1 unverified | report |
| 19 | Block-sparse LSTM | 94.99 | – | Paper | Code | 2017 | linked, not harvested | report |
| 20 | Space-XLNet | 94.88 | – | Paper | Code | 2024 | linked, not harvested | report |
| 21 | CEN-tpc | 94.52 | – | Paper | Code | 2017 | linked, not harvested | report |
| 22 | FLAN 137B (zero-shot) | 94.3 | ✓ | Paper | Code | 2021 | 0 of 1 ran · 1 unverified | report |
| 23 | oh-LSTM | 94.1 | ✓ | Paper | – | 2016 | no code linked | report |
| 24 | Virtual adversarial training | 94.1 | – | Paper | Code | 2016 | linked, not harvested | report |
| 25 | DV-ngrams-cosine + NB-weighted BON (re-evaluated) | 93.68 | – | Paper | Code | 2022 | linked, not harvested | report |
| 26 | Nyströmformer | 93.2 | – | Paper | Code | 2021 | 1 of 2 ran · 1 unverified | report |
| 27 | Modified LMU (34M) | 93.20 | – | Paper | Code | 2021 | linked, not harvested | report |
| 28 | DV-ngrams-cosine | 93.13 | – | Paper | Code | 2019 | linked, not harvested | report |
| 29 | OCaTS (kNN & GPT-3.5-turbo | 93.06 | – | Paper | Code | 2023 | 1 of 1 ran · 0 unverified | report |
| 30 | DistilBERT 66M | 92.82 | – | Paper | Code | 2019 | 19 of 27 ran · 8 unverified | report |
| 31 | GPT-2 Finetuned | 92.36 | – | Paper | Code | 2019 | linked, not harvested | report |
| 32 | seq2-bown-CNN | 92.33 | – | Paper | Code | 2014 | linked, not harvested | report |
| 33 | BP-Transformer + GloVe | 92.12 | – | Paper | Code | 2019 | 0 of 7 ran · 7 unverified | report |
| 34 | BCN+Char+CoVe | 91.8 | – | Paper | Code | 2017 | 3 of 3 ran · 0 unverified | report |
| 35 | ToWE-SG | 90.8 | – | Paper | Code | 2018 | linked, not harvested | report |
| 36 | COSINE | 90.54 | – | Paper | Code | 2020 | 0 of 5 ran · 5 unverified | report |
| 37 | LSTM with dynamic skip | 90.1 | – | Paper | Code | 2018 | linked, not harvested | report |
| 38 | CNN+LSTM | 88.9 | ✓ | Paper | Code | 2017 | linked, not harvested | report |
| 39 | UnICORNN | 88.4 | – | Paper | Code | 2021 | 1 of 1 ran · 0 unverified | report |
| 40 | CfC | 88.4 | – | Paper | Code | 2021 | 2 of 10 ran · 8 unverified | report |
| 41 | Doc2VecC | 88.3 | – | Paper | Code | 2017 | linked, not harvested | report |
| 42 | Bert+ Wilson-Cowan model RNN | 87.46 | – | Paper | Code | 2024 | linked, not harvested | report |
| 43 | coRNN | 87.4% | – | Paper | Code | 2020 | 1 of 3 ran · 2 unverified | report |
| 44 | S-LSTM | 87.15 | ✓ | Paper | Code | 2018 | 1 of 1 ran · 0 unverified | report |
| 45 | AlexNet [alexnet] | 87 | – | Paper | Code | 2021 | linked, not harvested | report |
| 46 | VGG16 [vgg16] | 86 | – | Paper | Code | 2021 | linked, not harvested | report |
| 47 | ResNext[resnext] | 85 | – | Paper | Code | 2021 | linked, not harvested | report |
| 48 | Standard DR-AGG | 45.1 | ✓ | Paper | Code | 2018 | linked, not harvested | report |
| 49 | Reverse DR-AGG | 44.5 | – | Paper | Code | 2018 | linked, not harvested | report |
All 49 rows shown. 48 link to a paper page on this site; 10 are marked as using additional training data in the archive. No GitHub stars are tracked; "Code" is the first repository the archive lists for the row. The archive carries no row tags, review links or community-submitted rows for this table; none are shown. archive 2025-07-28
Syntology Ran reads "N of M ran · U unverified": of the M code samples Syntology harvested from repositories linked to that row's paper (joined by arXiv id), N executed on a synthesized input and the other U = M−N are unverified (harvested, no recorded run). It counts code from repositories linked to that row's paper, not this result: the row's number was not reproduced and nothing here is a correctness claim. The other cell texts mean no graph line for the row: "linked, not harvested" (the archive links code, Syntology has not harvested it), "no code linked" (no code link in the archive), "not matched" (the row's paper URL matched no paper on this site). 21 rows have a graph line, from 18 distinct papers; 17 rows (15 papers) have at least one sample that ran. Counting each paper once: Syntology ran 66 of 172 samples; 106 unverified. Separately, 28 of those 172 are pointer-only (licence): the site points at that code rather than redistributing it, a licence property recorded for ran and unverified samples alike; each cell's tooltip carries the row's own pointer-only count. Read from the graph 2026-09-24. Per-sample status is on the paper page.
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