{"about":{"non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","site":"https://codewithpapers.app","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/single-layer-predictive-normalized-maximum","title":"Single Layer Predictive Normalized Maximum Likelihood for Out-of-Distribution Detection","arxiv_id":"2110.09246","date":"2021-10-18","proceeding":"NeurIPS 2021 12","authors":["Koby Bibas","Meir Feder","Tal Hassner"],"abstract":"Detecting out-of-distribution (OOD) samples is vital for developing machine learning based models for critical safety systems. Common approaches for OOD detection assume access to some OOD samples during training which may not be available in a real-life scenario. Instead, we utilize the {\\em predictive normalized maximum likelihood} (pNML) learner, in which no assumptions are made on the tested input. We derive an explicit expression of the pNML and its generalization error, denoted as the {\\em regret}, for a single layer neural network (NN). We show that this learner generalizes well when (i) the test vector resides in a subspace spanned by the eigenvectors associated with the large eigenvalues of the empirical correlation matrix of the training data, or (ii) the test sample is far from the decision boundary. Furthermore, we describe how to efficiently apply the derived pNML regret to any pretrained deep NN, by employing the explicit pNML for the last layer, followed by the softmax function. Applying the derived regret to deep NN requires neither additional tunable parameters nor extra data. We extensively evaluate our approach on 74 OOD detection benchmarks using DenseNet-100, ResNet-34, and WideResNet-40 models trained with CIFAR-100, CIFAR-10, SVHN, and ImageNet-30 showing a significant improvement of up to 15.6\\% over recent leading methods.","url_abs":"https://arxiv.org/abs/2110.09246v1","url_pdf":"https://arxiv.org/pdf/2110.09246v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"single-layer-predictive-normalized-maximum","repo_url":"https://github.com/kobybibas/pnml_ood_detection","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"ood-detection","task_name":"Out of Distribution (OOD) Detection"},{"task_slug":"out-of-distribution-detection","task_name":"Out-of-Distribution Detection"}],"methods":[{"method_slug":"test","method_name":"Test"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-8","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs Gaussian","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-8","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs Gaussian","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-6","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs ImageNet (C)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"99.9"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-6","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs ImageNet (C)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.8"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-5","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs ImageNet (R)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"99.9"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-5","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs ImageNet (R)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.9"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-3","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs LSUN (C)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"99.9"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-3","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs LSUN (C)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.5"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-9","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs LSUN (R)","model":"ResNet-34","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-9","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs LSUN (R)","model":"DenseNet-BC-100","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-2","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs SVHN","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":3,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-2","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs SVHN","model":"ResNet-34","rank_in_archive_order":3,"of":3,"metrics":{"AUROC":"99.8"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-7","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs Uniform","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-7","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs Uniform","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-4","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs iSUN","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-10-vs-4","task":"Out-of-Distribution Detection","dataset":"CIFAR-10 vs iSUN","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-7","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs Gaussian","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-7","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs Gaussian","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-5","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs ImageNet (C)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"99.0"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-5","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs ImageNet (C)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"98.4"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-4","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs ImageNet (R)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"99.5"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-4","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs ImageNet (R)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.2"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-2","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs LSUN (C)","model":"ResNet-34","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"97.8"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-2","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs LSUN (C)","model":"DenseNet-BC-100","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"96.1"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-3","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs LSUN (R)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"99.7"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-3","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs LSUN (R)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.6"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-8","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs SVHN","model":"DenseNet-BC-100","rank_in_archive_order":2,"of":5,"metrics":{"AUROC":"98.4"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-8","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs SVHN","model":"ResNet-34","rank_in_archive_order":3,"of":5,"metrics":{"AUROC":"97.9"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-6","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs Uniform","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-6","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs Uniform","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-1","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs iSUN","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"99.5"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-cifar-100-vs-1","task":"Out-of-Distribution Detection","dataset":"CIFAR-100 vs iSUN","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.3"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-4","task":"Out-of-Distribution Detection","dataset":"SVHN vs CIFAR-10","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-4","task":"Out-of-Distribution Detection","dataset":"SVHN vs CIFAR-10","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.8"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-5","task":"Out-of-Distribution Detection","dataset":"SVHN vs CIFAR-100","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-5","task":"Out-of-Distribution Detection","dataset":"SVHN vs CIFAR-100","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.8"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-3","task":"Out-of-Distribution Detection","dataset":"SVHN vs Gaussian","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-3","task":"Out-of-Distribution Detection","dataset":"SVHN vs Gaussian","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-1","task":"Out-of-Distribution Detection","dataset":"SVHN vs ImageNet (C)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-1","task":"Out-of-Distribution Detection","dataset":"SVHN vs ImageNet (C)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs","task":"Out-of-Distribution Detection","dataset":"SVHN vs ImageNet (R)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs","task":"Out-of-Distribution Detection","dataset":"SVHN vs ImageNet (R)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-lsun","task":"Out-of-Distribution Detection","dataset":"SVHN vs LSUN (C)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-lsun","task":"Out-of-Distribution Detection","dataset":"SVHN vs LSUN (C)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"99.9"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-lsun-1","task":"Out-of-Distribution Detection","dataset":"SVHN vs LSUN (R)","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-lsun-1","task":"Out-of-Distribution Detection","dataset":"SVHN vs LSUN (R)","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-2","task":"Out-of-Distribution Detection","dataset":"SVHN vs Uniform","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-2","task":"Out-of-Distribution Detection","dataset":"SVHN vs Uniform","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-isun","task":"Out-of-Distribution Detection","dataset":"SVHN vs iSUN","model":"DenseNet-BC-100","rank_in_archive_order":1,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false},{"leaderboard":"/sota/out-of-distribution-detection-on-svhn-vs-isun","task":"Out-of-Distribution Detection","dataset":"SVHN vs iSUN","model":"ResNet-34","rank_in_archive_order":2,"of":2,"metrics":{"AUROC":"100"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2110.09246","atlas_url":"https://app.syntology.ai/?focus=2110.09246","mcp":null,"developers":"https://syntology.ai/developers","held":true,"reason":"arxiv_id_in_atlas","checked_against":{"atlas_ids":549449,"atlas_ids_sha256":"afd72399dafd9906e9062927f70655952385a3d75488558e1f0484e950ab09b7","read_at":"2026-10-05T01:46:21Z"},"agent_calls":[{"tool":"get_citation_path","arguments":{"paper_1":"2110.09246"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"2110.09246\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"2110.09246"},"call":"get_concepts_for_paper(arxiv_id=\"2110.09246\")"}],"by_repo":[{"repo":"kobybibas/pnml_ood_detection","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":1,"harvested_for":["1610.02136"],"harvested_for_calls":[{"arxiv_id":"1610.02136","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1610.02136"},"call":"get_harvested_code_for_paper(arxiv_id=\"1610.02136\")"}]}],"by_repo_harvested_for_is":"harvested_for: the arXiv ids (as Syntology's graph holds them) of every paper whose samples Syntology's graph links to this repository, in arXiv-id order, complete; the page names the first 3 and says how many more. A link only: it does not say the papers are the same work, that one copied the other, or that the code implements this paper's method; harvested_for_calls gives the get_harvested_code_for_paper call for each arXiv-shaped id among them","by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code"},"code_links_note":null,"arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"a65a1ad53c57b00558ffbb41a92726032b394e8084c78399570bed99c415dcf1","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}