{"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/xifbench-evaluating-large-language-models-on","title":"XIFBench: Evaluating Large Language Models on Multilingual Instruction Following","arxiv_id":"2503.07539","date":"2025-03-10","proceeding":null,"authors":["Zhenyu Li","Kehai Chen","Yunfei Long","Xuefeng Bai","Yaoyin Zhang","Xuchen Wei","Juntao Li","Min Zhang"],"abstract":"Large Language Models (LLMs) have demonstrated remarkable instruction-following capabilities across various applications. However, their performance in multilingual settings remains poorly understood, as existing evaluations lack fine-grained constraint analysis. We introduce XIFBench, a comprehensive constraint-based benchmark for assessing multilingual instruction-following abilities of LLMs, featuring a novel taxonomy of five constraint categories and 465 parallel instructions across six languages spanning different resource levels. To ensure consistent cross-lingual evaluation, we develop a requirement-based protocol that leverages English requirements as semantic anchors. These requirements are then used to validate the translations across languages. Extensive experiments with various LLMs reveal notable variations in instruction-following performance across resource levels, identifying key influencing factors such as constraint categories, instruction complexity, and cultural specificity.","url_abs":"https://arxiv.org/abs/2503.07539v1","url_pdf":"https://arxiv.org/pdf/2503.07539v1.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":[],"tasks":[{"task_slug":"instruction-following","task_name":"Instruction Following"},{"task_slug":"specificity","task_name":"Specificity"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2503.07539","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2503.07539"}},"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":"deterministic:regex_extraction","url":"https://github.com/zhenyuli801/XIFBench","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":12},"by_repo_kind":{"found_in_text":{"samples":12,"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":"3a43b74a35f656eb","entry":"format_evaluation_requirements","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"D2_validate_translations.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/D2_validate_translations.py","link_basis":"harvester_set","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":"3a43b74a35f656eb"}},{"code_sha256_prefix":"1cfb172de71361fd","entry":"format_sampled_constraints","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"B2_combine_constraints.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/B2_combine_constraints.py","link_basis":"harvester_set","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":"1cfb172de71361fd"}},{"code_sha256_prefix":"635301a89612a3d4","entry":"generate_reasoning_and_responses","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"G1_api_inference.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/G1_api_inference.py","link_basis":"harvester_set","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":"635301a89612a3d4"}},{"code_sha256_prefix":"d52e4d9cbd07b9c1","entry":"get_openai_response","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"B1_brainstorm_constraints.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/B1_brainstorm_constraints.py","link_basis":"harvester_set","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":"d52e4d9cbd07b9c1"}},{"code_sha256_prefix":"1aac49f7aa943d77","entry":"load_alpacaeval","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"A1_sample_instructions.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/A1_sample_instructions.py","link_basis":"harvester_set","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":"1aac49f7aa943d77"}},{"code_sha256_prefix":"cf254803acaaf54c","entry":"load_json","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"A1_sample_instructions.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/A1_sample_instructions.py","link_basis":"harvester_set","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":"cf254803acaaf54c"}},{"code_sha256_prefix":"d4c04971c3f6dc69","entry":"load_jsonl","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"A1_sample_instructions.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/A1_sample_instructions.py","link_basis":"harvester_set","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":"d4c04971c3f6dc69"}},{"code_sha256_prefix":"c104e11d10734d8e","entry":"parse_constraint_brainstorming_response","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"B1_brainstorm_constraints.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/B1_brainstorm_constraints.py","link_basis":"harvester_set","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":"c104e11d10734d8e"}},{"code_sha256_prefix":"e02d2d2f5e843460","entry":"parse_constraint_combination_response","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"B2_combine_constraints.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/B2_combine_constraints.py","link_basis":"harvester_set","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":"e02d2d2f5e843460"}},{"code_sha256_prefix":"505083d5b2dec3e6","entry":"parse_requirement_categorization_response","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"C2_categorize_requirements.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/C2_categorize_requirements.py","link_basis":"harvester_set","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":"505083d5b2dec3e6"}},{"code_sha256_prefix":"d8bad0118c48538d","entry":"parse_requirement_decomposition_response","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"C1_decompose_requirements.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/C1_decompose_requirements.py","link_basis":"harvester_set","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":"d8bad0118c48538d"}},{"code_sha256_prefix":"33730eed64c4e455","entry":"sample_constraints","repo":"zhenyuli801/XIFBench","repo_kind":"found_in_text","path":"B2_combine_constraints.py","file_url":"https://github.com/zhenyuli801/XIFBench/blob/HEAD/B2_combine_constraints.py","link_basis":"harvester_set","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":"33730eed64c4e455"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}