{"about":{"site":"https://codewithpapers.app","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.","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"},"url":"/paper/do-text-free-diffusion-models-learn","title":"Do text-free diffusion models learn discriminative visual representations?","arxiv_id":"2311.17921","date":"2023-11-29","proceeding":null,"authors":["Soumik Mukhopadhyay","Matthew Gwilliam","Yosuke Yamaguchi","Vatsal Agarwal","Namitha Padmanabhan","Archana Swaminathan","Tianyi Zhou","Jun Ohya","Abhinav Shrivastava"],"abstract":"While many unsupervised learning models focus on one family of tasks, either generative or discriminative, we explore the possibility of a unified representation learner: a model which addresses both families of tasks simultaneously. We identify diffusion models, a state-of-the-art method for generative tasks, as a prime candidate. Such models involve training a U-Net to iteratively predict and remove noise, and the resulting model can synthesize high-fidelity, diverse, novel images. We find that the intermediate feature maps of the U-Net are diverse, discriminative feature representations. We propose a novel attention mechanism for pooling feature maps and further leverage this mechanism as DifFormer, a transformer feature fusion of features from different diffusion U-Net blocks and noise steps. We also develop DifFeed, a novel feedback mechanism tailored to diffusion. We find that diffusion models are better than GANs, and, with our fusion and feedback mechanisms, can compete with state-of-the-art unsupervised image representation learning methods for discriminative tasks - image classification with full and semi-supervision, transfer for fine-grained classification, object detection and segmentation, and semantic segmentation. Our project website (https://mgwillia.github.io/diffssl/) and code (https://github.com/soumik-kanad/diffssl) are available publicly.","url_abs":"https://arxiv.org/abs/2311.17921v3","url_pdf":"https://arxiv.org/pdf/2311.17921v3.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":"do-text-free-diffusion-models-learn","repo_url":"https://github.com/soumik-kanad/diffssl","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"image-classification","task_name":"image-classification"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[{"method_slug":"concatenated-skip-connection","method_name":"Concatenated Skip Connection"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"diffusion","method_name":"Diffusion"},{"method_slug":"focus","method_name":"Focus"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"u-net","method_name":"U-Net"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2311.17921","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2311.17921"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/soumik-kanad/diffssl","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":2,"ran_honours":3,"ran":5,"unverified":4},"by_repo_kind":{"official":{"samples":14,"ran":10,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"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"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"cfd76fd0d89574a4","entry":"approx_standard_normal_cdf","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/losses.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/losses.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cfd76fd0d89574a4"}},{"code_sha256_prefix":"2ab2316ac6fdd869","entry":"betas_for_alpha_bar","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/gaussian_diffusion.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/gaussian_diffusion.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2ab2316ac6fdd869"}},{"code_sha256_prefix":"cd33283d615fb3d7","entry":"discretized_gaussian_log_likelihood","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/losses.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/losses.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cd33283d615fb3d7"}},{"code_sha256_prefix":"a086d6286a40b889","entry":"get_named_beta_schedule","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/gaussian_diffusion.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/gaussian_diffusion.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a086d6286a40b889"}},{"code_sha256_prefix":"e41367ad14ff58fd","entry":"get_param_groups_and_shapes","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/fp16_util.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/fp16_util.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e41367ad14ff58fd"}},{"code_sha256_prefix":"e20dd5102da3b050","entry":"make_master_params","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/fp16_util.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/fp16_util.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e20dd5102da3b050"}},{"code_sha256_prefix":"c4b9f60ffc5a8bac","entry":"make_output_format","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/logger.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/logger.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c4b9f60ffc5a8bac"}},{"code_sha256_prefix":"cf2798b666b231ca","entry":"normal_kl","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/losses.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/losses.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cf2798b666b231ca"}},{"code_sha256_prefix":"4745538bfc32f1f9","entry":"random_crop_arr","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/image_datasets.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/image_datasets.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4745538bfc32f1f9"}},{"code_sha256_prefix":"64fff1e30802b815","entry":"unflatten_master_params","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/fp16_util.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/fp16_util.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"64fff1e30802b815"}},{"code_sha256_prefix":"f8b4a29a52612a41","entry":"center_crop_arr","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/image_datasets.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/image_datasets.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f8b4a29a52612a41"}},{"code_sha256_prefix":"e515a67f7f32e76d","entry":"mpi_weighted_mean","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/logger.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/logger.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e515a67f7f32e76d"}},{"code_sha256_prefix":"79d5420720b77480","entry":"profile","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"guided_diffusion/logger.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/guided_diffusion/logger.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"79d5420720b77480"}},{"code_sha256_prefix":"07badd81261def9b","entry":"sync_tensor_across_gpus","repo":"soumik-kanad/diffssl","repo_kind":"official","path":"finetune.py","file_url":"https://github.com/soumik-kanad/diffssl/blob/HEAD/finetune.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"07badd81261def9b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}