{"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/a-closer-look-at-the-training-dynamics-of","title":"Understanding the Role of the Projector in Knowledge Distillation","arxiv_id":"2303.11098","date":"2023-03-20","proceeding":null,"authors":["Roy Miles","Krystian Mikolajczyk"],"abstract":"In this paper we revisit the efficacy of knowledge distillation as a function matching and metric learning problem. In doing so we verify three important design decisions, namely the normalisation, soft maximum function, and projection layers as key ingredients. We theoretically show that the projector implicitly encodes information on past examples, enabling relational gradients for the student. We then show that the normalisation of representations is tightly coupled with the training dynamics of this projector, which can have a large impact on the students performance. Finally, we show that a simple soft maximum function can be used to address any significant capacity gap problems. Experimental results on various benchmark datasets demonstrate that using these insights can lead to superior or comparable performance to state-of-the-art knowledge distillation techniques, despite being much more computationally efficient. In particular, we obtain these results across image classification (CIFAR100 and ImageNet), object detection (COCO2017), and on more difficult distillation objectives, such as training data efficient transformers, whereby we attain a 77.2% top-1 accuracy with DeiT-Ti on ImageNet. Code and models are publicly available.","url_abs":"https://arxiv.org/abs/2303.11098v5","url_pdf":"https://arxiv.org/pdf/2303.11098v5.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":"a-closer-look-at-the-training-dynamics-of","repo_url":"https://github.com/roymiles/simple-recipe-distillation","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"a-closer-look-at-the-training-dynamics-of","repo_url":"https://github.com/roymiles/vkd","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"a-closer-look-at-the-training-dynamics-of","repo_url":"https://github.com/yoshitomo-matsubara/torchdistill","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"a-closer-look-at-the-training-dynamics-of","repo_url":"https://github.com/Hazqeel09/ellzaf_ml","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"knowledge-distillation","task_name":"Knowledge Distillation"},{"task_slug":"metric-learning","task_name":"Metric Learning"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"image-classification","task_name":"image-classification"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[{"method_slug":"knowledge-distillation","method_name":"Knowledge Distillation"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/knowledge-distillation-on-cifar-100","task":"Knowledge Distillation","dataset":"CIFAR-100","model":"SRD (T:resnet-32x4, S:shufflenet-v2)","rank_in_archive_order":1,"of":27,"metrics":{"Top-1 Accuracy (%)":"79.86"},"uses_additional_data":false},{"leaderboard":"/sota/knowledge-distillation-on-imagenet","task":"Knowledge Distillation","dataset":"ImageNet","model":"SRD (T:RegNety 160 S:DeiT-S)","rank_in_archive_order":12,"of":52,"metrics":{"CRD training setting":"✘","Top-1 accuracy %":"82.1","model size":"22M"},"uses_additional_data":false},{"leaderboard":"/sota/knowledge-distillation-on-imagenet","task":"Knowledge Distillation","dataset":"ImageNet","model":"SRD (T:RegNety 160 S:DeIT-Ti)","rank_in_archive_order":21,"of":52,"metrics":{"CRD training setting":"✘","Top-1 accuracy %":"77.2","model size":"6M"},"uses_additional_data":false},{"leaderboard":"/sota/knowledge-distillation-on-imagenet","task":"Knowledge Distillation","dataset":"ImageNet","model":"SRD (T: ResNet-34 S:ResNet-18)","rank_in_archive_order":38,"of":52,"metrics":{"CRD training setting":"✓","Top-1 accuracy %":"71.87"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2303.11098","atlas_url":"https://app.syntology.ai/?focus=2303.11098","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":"2303.11098"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"2303.11098\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"2303.11098"},"call":"get_concepts_for_paper(arxiv_id=\"2303.11098\")"}],"by_repo":[{"repo":"roymiles/simple-recipe-distillation","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0},{"repo":"roymiles/vkd","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":1,"harvested_for":["2403.06213"],"harvested_for_calls":[{"arxiv_id":"2403.06213","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2403.06213"},"call":"get_harvested_code_for_paper(arxiv_id=\"2403.06213\")"}]},{"repo":"yoshitomo-matsubara/torchdistill","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":6,"harvested_for":["1503.02531","1512.03385","1605.07146","2011.12913","2104.09044","2205.10536"],"harvested_for_calls":[{"arxiv_id":"1503.02531","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1503.02531"},"call":"get_harvested_code_for_paper(arxiv_id=\"1503.02531\")"},{"arxiv_id":"1512.03385","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1512.03385"},"call":"get_harvested_code_for_paper(arxiv_id=\"1512.03385\")"},{"arxiv_id":"1605.07146","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1605.07146"},"call":"get_harvested_code_for_paper(arxiv_id=\"1605.07146\")"},{"arxiv_id":"2011.12913","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2011.12913"},"call":"get_harvested_code_for_paper(arxiv_id=\"2011.12913\")"},{"arxiv_id":"2104.09044","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2104.09044"},"call":"get_harvested_code_for_paper(arxiv_id=\"2104.09044\")"},{"arxiv_id":"2205.10536","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2205.10536"},"call":"get_harvested_code_for_paper(arxiv_id=\"2205.10536\")"}]},{"repo":"Hazqeel09/ellzaf_ml","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":2,"harvested_for":["2111.09886","2304.06446"],"harvested_for_calls":[{"arxiv_id":"2111.09886","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2111.09886"},"call":"get_harvested_code_for_paper(arxiv_id=\"2111.09886\")"},{"arxiv_id":"2304.06446","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2304.06446"},"call":"get_harvested_code_for_paper(arxiv_id=\"2304.06446\")"}]}],"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":"3cfac9d6e981be18a9b5dd3ec7585c04eb9e456d76f918825838956cebc9a581","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}