{"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/m-m-s-a-benchmark-to-evaluate-tool-use-for","title":"m&m's: A Benchmark to Evaluate Tool-Use for multi-step multi-modal Tasks","arxiv_id":"2403.11085","date":"2024-03-17","proceeding":null,"authors":["Zixian Ma","Weikai Huang","Jieyu Zhang","Tanmay Gupta","Ranjay Krishna"],"abstract":"Real-world multi-modal problems are rarely solved by a single machine learning model, and often require multi-step computational plans that involve stitching several models. Tool-augmented LLMs hold tremendous promise for automating the generation of such computational plans. However, the lack of standardized benchmarks for evaluating LLMs as planners for multi-step multi-modal tasks has prevented a systematic study of planner design decisions. Should LLMs generate a full plan in a single shot or step-by-step? Should they invoke tools directly with Python code or through structured data formats like JSON? Does feedback improve planning? To answer these questions and more, we introduce m&m's: a benchmark containing 4K+ multi-step multi-modal tasks involving 33 tools that include multi-modal models, (free) public APIs, and image processing modules. For each of these task queries, we provide automatically generated plans using this realistic toolset. We further provide a high-quality subset of 1,565 task plans that are human-verified and correctly executable. With m&m's, we evaluate 10 popular LLMs with 2 planning strategies (multi-step vs. step-by-step planning), 2 plan formats (JSON vs. code), and 3 types of feedback (parsing/verification/execution). Finally, we summarize takeaways from our extensive experiments. Our dataset and code are available on HuggingFace (https://huggingface.co/datasets/zixianma/mnms) and Github (https://github.com/RAIVNLab/mnms).","url_abs":"https://arxiv.org/abs/2403.11085v4","url_pdf":"https://arxiv.org/pdf/2403.11085v4.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":"m-m-s-a-benchmark-to-evaluate-tool-use-for","repo_url":"https://github.com/raivnlab/mnms","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"m-m-s-a-benchmark-to-evaluate-tool-use-for","repo_url":"https://github.com/RAIVNLab/mms","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"4k","task_name":"4k"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2403.11085","atlas_url":"https://app.syntology.ai/?focus=2403.11085","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2403.11085"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/raivnlab/mnms","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/RAIVNLab/mms","reach":{"status":"ok"}}],"summary":{"ran_violates":1,"ran":7,"ran_draft_wrong":1,"unverified":2},"by_repo_kind":{"official":{"samples":11,"ran":9,"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":11,"samples":[{"code_sha256_prefix":"9c26cc8e53781756","entry":"checks_terminate_message","repo":"raivnlab/mnms","repo_kind":"official","path":"mnms/run_plan_agent.py","file_url":"https://github.com/raivnlab/mnms/blob/HEAD/mnms/run_plan_agent.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9c26cc8e53781756"}},{"code_sha256_prefix":"cd95a2581690e761","entry":"find_node_patterns","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/verifier.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/verifier.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"cd95a2581690e761"}},{"code_sha256_prefix":"bf0b50da342cb58e","entry":"flatten_labels","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/evaluation/metrics.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/evaluation/metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"bf0b50da342cb58e"}},{"code_sha256_prefix":"6bb1d852d31d4d01","entry":"format_python_str","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/prompt.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/prompt.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"6bb1d852d31d4d01"}},{"code_sha256_prefix":"1a468c0ba7ba78eb","entry":"nodes_to_code","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/evaluation/evaluator.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/evaluation/evaluator.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"1a468c0ba7ba78eb"}},{"code_sha256_prefix":"5e9856951ad7a58e","entry":"nodes_to_program","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/prompt.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/prompt.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"5e9856951ad7a58e"}},{"code_sha256_prefix":"cd58034ecc1945fe","entry":"object_processing","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/execution/tool_api.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/execution/tool_api.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"cd58034ecc1945fe"}},{"code_sha256_prefix":"5822609079590d9e","entry":"read_data_from_file","repo":"raivnlab/mnms","repo_kind":"official","path":"mnms/run_plan_agent.py","file_url":"https://github.com/raivnlab/mnms/blob/HEAD/mnms/run_plan_agent.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"5822609079590d9e"}},{"code_sha256_prefix":"76350c7f8fa4a0c6","entry":"replace_node_with_placeholder","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/evaluation/evaluator.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/evaluation/evaluator.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"76350c7f8fa4a0c6"}},{"code_sha256_prefix":"0776d0fd85848acd","entry":"find_best_match","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/evaluation/evaluator.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/evaluation/evaluator.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"0776d0fd85848acd"}},{"code_sha256_prefix":"f7710cfb5c0415cb","entry":"get_full_path_data","repo":"RAIVNLab/mms","repo_kind":"official","path":"mnms/execution/tool_api.py","file_url":"https://github.com/RAIVNLab/mms/blob/HEAD/mnms/execution/tool_api.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f7710cfb5c0415cb"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}