{"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-study-of-autoregressive-decoders-for-multi","title":"A Study of Autoregressive Decoders for Multi-Tasking in Computer Vision","arxiv_id":"2303.17376","date":"2023-03-30","proceeding":null,"authors":["Lucas Beyer","Bo Wan","Gagan Madan","Filip Pavetic","Andreas Steiner","Alexander Kolesnikov","André Susano Pinto","Emanuele Bugliarello","Xiao Wang","Qihang Yu","Liang-Chieh Chen","Xiaohua Zhai"],"abstract":"There has been a recent explosion of computer vision models which perform many tasks and are composed of an image encoder (usually a ViT) and an autoregressive decoder (usually a Transformer). However, most of this work simply presents one system and its results, leaving many questions regarding design decisions and trade-offs of such systems unanswered. In this work, we aim to provide such answers. We take a close look at autoregressive decoders for multi-task learning in multimodal computer vision, including classification, captioning, visual question answering, and optical character recognition. Through extensive systematic experiments, we study the effects of task and data mixture, training and regularization hyperparameters, conditioning type and specificity, modality combination, and more. Importantly, we compare these to well-tuned single-task baselines to highlight the cost incurred by multi-tasking. A key finding is that a small decoder learned on top of a frozen pretrained encoder works surprisingly well. We call this setup locked-image tuning with decoder (LiT-decoder). It can be seen as teaching a decoder to interact with a pretrained vision model via natural language.","url_abs":"https://arxiv.org/abs/2303.17376v1","url_pdf":"https://arxiv.org/pdf/2303.17376v1.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-study-of-autoregressive-decoders-for-multi","repo_url":"https://github.com/google-research/big_vision","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"jax","reach":null}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"},{"task_slug":"optical-character-recognition","task_name":"Optical Character Recognition"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"specificity","task_name":"Specificity"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2303.17376","atlas_url":"https://app.syntology.ai/?focus=2303.17376","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":"2303.17376"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"2303.17376\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"2303.17376"},"call":"get_concepts_for_paper(arxiv_id=\"2303.17376\")"}],"by_repo":[{"repo":"google-research/big_vision","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":7,"harvested_for":["2106.10270","2205.10337","2209.06794","2305.16999","2306.07915","2407.07726","2411.19722"],"harvested_for_calls":[{"arxiv_id":"2106.10270","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2106.10270"},"call":"get_harvested_code_for_paper(arxiv_id=\"2106.10270\")"},{"arxiv_id":"2205.10337","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2205.10337"},"call":"get_harvested_code_for_paper(arxiv_id=\"2205.10337\")"},{"arxiv_id":"2209.06794","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2209.06794"},"call":"get_harvested_code_for_paper(arxiv_id=\"2209.06794\")"},{"arxiv_id":"2305.16999","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2305.16999"},"call":"get_harvested_code_for_paper(arxiv_id=\"2305.16999\")"},{"arxiv_id":"2306.07915","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2306.07915"},"call":"get_harvested_code_for_paper(arxiv_id=\"2306.07915\")"},{"arxiv_id":"2407.07726","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2407.07726"},"call":"get_harvested_code_for_paper(arxiv_id=\"2407.07726\")"},{"arxiv_id":"2411.19722","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2411.19722"},"call":"get_harvested_code_for_paper(arxiv_id=\"2411.19722\")"}]}],"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":"b02335ddb1b7cd9c5c08cc6b2cfc6dbb254d30ee4aa39a84cf7487cc6ebd2c71","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}