{"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/waterbench-towards-holistic-evaluation-of","title":"WaterBench: Towards Holistic Evaluation of Watermarks for Large Language Models","arxiv_id":"2311.07138","date":"2023-11-13","proceeding":null,"authors":["Shangqing Tu","Yuliang Sun","Yushi Bai","Jifan Yu","Lei Hou","Juanzi Li"],"abstract":"To mitigate the potential misuse of large language models (LLMs), recent research has developed watermarking algorithms, which restrict the generation process to leave an invisible trace for watermark detection. Due to the two-stage nature of the task, most studies evaluate the generation and detection separately, thereby presenting a challenge in unbiased, thorough, and applicable evaluations. In this paper, we introduce WaterBench, the first comprehensive benchmark for LLM watermarks, in which we design three crucial factors: (1) For benchmarking procedure, to ensure an apples-to-apples comparison, we first adjust each watermarking method's hyper-parameter to reach the same watermarking strength, then jointly evaluate their generation and detection performance. (2) For task selection, we diversify the input and output length to form a five-category taxonomy, covering $9$ tasks. (3) For evaluation metric, we adopt the GPT4-Judge for automatically evaluating the decline of instruction-following abilities after watermarking. We evaluate $4$ open-source watermarks on $2$ LLMs under $2$ watermarking strengths and observe the common struggles for current methods on maintaining the generation quality. The code and data are available at https://github.com/THU-KEG/WaterBench.","url_abs":"https://arxiv.org/abs/2311.07138v2","url_pdf":"https://arxiv.org/pdf/2311.07138v2.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":"waterbench-towards-holistic-evaluation-of","repo_url":"https://github.com/THU-KEG/WaterBench","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null},{"paper_slug":"waterbench-towards-holistic-evaluation-of","repo_url":"https://github.com/Xieyangxinyu/Unbiased-Watermark-via-Maximal-Coupling","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"waterbench-towards-holistic-evaluation-of","repo_url":"https://github.com/dortsur/heavywater_simplexwater","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"benchmarking","task_name":"Benchmarking"},{"task_slug":"instruction-following","task_name":"Instruction Following"}],"methods":[],"datasets_introduced":[{"slug":"waterbench","name":"WaterBench","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2311.07138","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2311.07138"}},"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/Xieyangxinyu/Unbiased-Watermark-via-Maximal-Coupling","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/THU-KEG/WaterBench","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/dortsur/heavywater_simplexwater","reach":null}],"summary":{"ran_fixture":1,"ran_violates":2,"ran_draft_wrong":1,"unverified":2},"by_repo_kind":{"official":{"samples":2,"ran":1,"repositories":1},"listed":{"samples":3,"ran":2,"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":1,"samples":[{"code_sha256_prefix":"e50c48cad3cee475","entry":"bootstrap_pd","repo":"dortsur/heavywater_simplexwater","repo_kind":"listed","path":"detect.py","file_url":"https://github.com/dortsur/heavywater_simplexwater/blob/HEAD/detect.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e50c48cad3cee475"}},{"code_sha256_prefix":"ed4be202f208766c","entry":"build_chat","repo":"THU-KEG/WaterBench","repo_kind":"official","path":"pred.py","file_url":"https://github.com/THU-KEG/WaterBench/blob/HEAD/pred.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ed4be202f208766c"}},{"code_sha256_prefix":"f59d3f952a8d5950","entry":"build_chat","repo":"dortsur/heavywater_simplexwater","repo_kind":"listed","path":"pred.py","file_url":"https://github.com/dortsur/heavywater_simplexwater/blob/HEAD/pred.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":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f59d3f952a8d5950"}},{"code_sha256_prefix":"678874b625786116","entry":"str2bool","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"678874b625786116"}},{"code_sha256_prefix":"a37634f5d95e4bc6","entry":"parse_args","repo":"THU-KEG/WaterBench","repo_kind":"official","path":"pred.py","file_url":"https://github.com/THU-KEG/WaterBench/blob/HEAD/pred.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":"a37634f5d95e4bc6"}},{"code_sha256_prefix":"07264568683ce46b","entry":"parse_args","repo":"dortsur/heavywater_simplexwater","repo_kind":"listed","path":"pred.py","file_url":"https://github.com/dortsur/heavywater_simplexwater/blob/HEAD/pred.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":"07264568683ce46b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}