{"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/gradient-free-activation-maximization-for","title":"Gradient-free activation maximization for identifying effective stimuli","arxiv_id":"1905.00378","date":"2019-05-01","proceeding":null,"authors":["Will Xiao","Gabriel Kreiman"],"abstract":"A fundamental question for understanding brain function is what types of\nstimuli drive neurons to fire. In visual neuroscience, this question has also\nbeen posted as characterizing the receptive field of a neuron. The search for\neffective stimuli has traditionally been based on a combination of insights\nfrom previous studies, intuition, and luck. Recently, the same question has\nemerged in the study of units in convolutional neural networks (ConvNets), and\ntogether with this question a family of solutions were developed that are\ngenerally referred to as \"feature visualization by activation maximization.\"\n  We sought to bring in tools and techniques developed for studying ConvNets to\nthe study of biological neural networks. However, one key difference that\nimpedes direct translation of tools is that gradients can be obtained from\nConvNets using backpropagation, but such gradients are not available from the\nbrain. To circumvent this problem, we developed a method for gradient-free\nactivation maximization by combining a generative neural network with a genetic\nalgorithm. We termed this method XDream (EXtending DeepDream with real-time\nevolution for activation maximization), and we have shown that this method can\nreliably create strong stimuli for neurons in the macaque visual cortex (Ponce\net al., 2019). In this paper, we describe extensive experiments characterizing\nthe XDream method by using ConvNet units as in silico models of neurons. We\nshow that XDream is applicable across network layers, architectures, and\ntraining sets; examine design choices in the algorithm; and provide practical\nguides for choosing hyperparameters in the algorithm. XDream is an efficient\nalgorithm for uncovering neuronal tuning preferences in black-box networks\nusing a vast and diverse stimulus space.","url_abs":"http://arxiv.org/abs/1905.00378v1","url_pdf":"http://arxiv.org/pdf/1905.00378v1.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":"gradient-free-activation-maximization-for","repo_url":"https://github.com/willwx/XDream","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1905.00378","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1905.00378"}},"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/willwx/XDream","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":7},"by_repo_kind":{"official":{"samples":7,"ran":0,"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":"5f288d9caa6dfb91","entry":"get_paths","repo":"willwx/XDream","repo_kind":"official","path":"xdream/net_utils/net_loader.py","file_url":"https://github.com/willwx/XDream/blob/HEAD/xdream/net_utils/net_loader.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":"5f288d9caa6dfb91"}},{"code_sha256_prefix":"9664d8c4c76bbc63","entry":"get_transformer","repo":"willwx/XDream","repo_kind":"official","path":"xdream/net_utils/net_loader.py","file_url":"https://github.com/willwx/XDream/blob/HEAD/xdream/net_utils/net_loader.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":"9664d8c4c76bbc63"}},{"code_sha256_prefix":"97faecd17bf0365b","entry":"load_codes","repo":"willwx/XDream","repo_kind":"official","path":"xdream/utils.py","file_url":"https://github.com/willwx/XDream/blob/HEAD/xdream/utils.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":"97faecd17bf0365b"}},{"code_sha256_prefix":"e98d5e11ca90b8c4","entry":"load_net","repo":"willwx/XDream","repo_kind":"official","path":"xdream/net_utils/net_loader.py","file_url":"https://github.com/willwx/XDream/blob/HEAD/xdream/net_utils/net_loader.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":"e98d5e11ca90b8c4"}},{"code_sha256_prefix":"9468bedf7c0fb9f4","entry":"read_image","repo":"willwx/XDream","repo_kind":"official","path":"xdream/utils.py","file_url":"https://github.com/willwx/XDream/blob/HEAD/xdream/utils.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":"9468bedf7c0fb9f4"}},{"code_sha256_prefix":"74848d8128c1942e","entry":"resize_image","repo":"willwx/XDream","repo_kind":"official","path":"xdream/net_utils/transformer.py","file_url":"https://github.com/willwx/XDream/blob/HEAD/xdream/net_utils/transformer.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":"74848d8128c1942e"}},{"code_sha256_prefix":"544047023db0e72f","entry":"resize_image","repo":"willwx/XDream","repo_kind":"official","path":"xdream/utils.py","file_url":"https://github.com/willwx/XDream/blob/HEAD/xdream/utils.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":"544047023db0e72f"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}