{"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/scaling-proprioceptive-visual-learning-with","title":"Scaling Proprioceptive-Visual Learning with Heterogeneous Pre-trained Transformers","arxiv_id":"2409.20537","date":"2024-09-30","proceeding":null,"authors":["Lirui Wang","Xinlei Chen","Jialiang Zhao","Kaiming He"],"abstract":"One of the roadblocks for training generalist robotic models today is heterogeneity. Previous robot learning methods often collect data to train with one specific embodiment for one task, which is expensive and prone to overfitting. This work studies the problem of learning policy representations through heterogeneous pre-training on robot data across different embodiments and tasks at scale. We propose Heterogeneous Pre-trained Transformers (HPT), which pre-train a large, shareable trunk of a policy neural network to learn a task and embodiment agnostic shared representation. This general architecture aligns the specific proprioception and vision inputs from distinct embodiments to a short sequence of tokens and then processes such tokens to map to control robots for different tasks. Leveraging the recent large-scale multi-embodiment real-world robotic datasets as well as simulation, deployed robots, and human video datasets, we investigate pre-training policies across heterogeneity. We conduct experiments to investigate the scaling behaviors of training objectives, to the extent of 52 datasets. HPTs outperform several baselines and enhance the fine-tuned policy performance by over 20% on unseen tasks in multiple simulator benchmarks and real-world settings. See the project website (https://liruiw.github.io/hpt/) for code and videos.","url_abs":"https://arxiv.org/abs/2409.20537v1","url_pdf":"https://arxiv.org/pdf/2409.20537v1.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":"scaling-proprioceptive-visual-learning-with","repo_url":"https://github.com/liruiw/HPT","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"scaling-proprioceptive-visual-learning-with","repo_url":"https://github.com/liruiw/lerobot","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2409.20537","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2409.20537"}},"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/liruiw/lerobot","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/liruiw/HPT","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":10,"ran_honours":1,"unverified":2},"by_repo_kind":{"official":{"samples":13,"ran":11,"repositories":2}},"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":"a3df0b772433ba60","entry":"add_noise","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/data_aug.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/data_aug.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a3df0b772433ba60"}},{"code_sha256_prefix":"74a689ae7347ab41","entry":"add_noise_depth","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/data_aug.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/data_aug.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"74a689ae7347ab41"}},{"code_sha256_prefix":"cf6b4fee94b1e3d7","entry":"chromatic_transform","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/data_aug.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/data_aug.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":"cf6b4fee94b1e3d7"}},{"code_sha256_prefix":"2b357622cfdee80a","entry":"euler2mat_batch","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/common_utils.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/common_utils.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":"2b357622cfdee80a"}},{"code_sha256_prefix":"9b7d923de2c72b94","entry":"get_sinusoid_encoding_table","repo":"liruiw/lerobot","repo_kind":"official","path":"lerobot/common/policies/hpt/modeling_hpt.py","file_url":"https://github.com/liruiw/lerobot/blob/HEAD/lerobot/common/policies/hpt/modeling_hpt.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"9b7d923de2c72b94"}},{"code_sha256_prefix":"5f2aa450516e0961","entry":"mat2euler_batch","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/common_utils.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/common_utils.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":"5f2aa450516e0961"}},{"code_sha256_prefix":"4da54929f5e00515","entry":"module_max_gradient","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/model_utils.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/model_utils.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":"4da54929f5e00515"}},{"code_sha256_prefix":"00a047267cd8287f","entry":"module_max_param","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/model_utils.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/model_utils.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":"00a047267cd8287f"}},{"code_sha256_prefix":"de6c132a5e226f00","entry":"module_mean_param","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/model_utils.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/model_utils.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":"de6c132a5e226f00"}},{"code_sha256_prefix":"7eff256e42a39e9f","entry":"select_image","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/dataset/local_traj_dataset.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/dataset/local_traj_dataset.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":"7eff256e42a39e9f"}},{"code_sha256_prefix":"319fcc44fd3c161c","entry":"select_proprioception","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/dataset/local_traj_dataset.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/dataset/local_traj_dataset.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":"319fcc44fd3c161c"}},{"code_sha256_prefix":"750875069e50bdd3","entry":"euler2axangle_batch","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/utils/common_utils.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/utils/common_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":"750875069e50bdd3"}},{"code_sha256_prefix":"3fd39d250788bb6c","entry":"vit_base_patch16","repo":"liruiw/HPT","repo_kind":"official","path":"hpt/models/policy_stem.py","file_url":"https://github.com/liruiw/HPT/blob/HEAD/hpt/models/policy_stem.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":"3fd39d250788bb6c"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}