{"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/agentbreeder-mitigating-the-ai-safety-impact","title":"AgentBreeder: Mitigating the AI Safety Impact of Multi-Agent Scaffolds via Self-Improvement","arxiv_id":"2502.00757","date":"2025-02-02","proceeding":null,"authors":["J Rosser","Jakob Nicolaus Foerster"],"abstract":"Scaffolding Large Language Models (LLMs) into multi-agent systems often improves performance on complex tasks, but the safety impact of such scaffolds has not been thoroughly explored. We introduce AgentBreeder, a framework for multi-objective self-improving evolutionary search over scaffolds. We evaluate discovered scaffolds on widely recognized reasoning, mathematics, and safety benchmarks and compare them with popular baselines. In 'blue' mode, we see a 79.4% average uplift in safety benchmark performance while maintaining or improving capability scores. In 'red' mode, we find adversarially weak scaffolds emerging concurrently with capability optimization. Our work demonstrates the risks of multi-agent scaffolding and provides a framework for mitigating them. Code is available at https://github.com/J-Rosser-UK/AgentBreeder.","url_abs":"https://arxiv.org/abs/2502.00757v3","url_pdf":"https://arxiv.org/pdf/2502.00757v3.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":"agentbreeder-mitigating-the-ai-safety-impact","repo_url":"https://github.com/J-Rosser-UK/AgentBreeder","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[{"method_slug":"base","method_name":"BASE"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2502.00757","atlas_url":"https://app.syntology.ai/?focus=2502.00757","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2502.00757"}},"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/J-Rosser-UK/AgentBreeder","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":2,"unverified":1},"by_repo_kind":{"official":{"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":0,"samples":[{"code_sha256_prefix":"f8bdf17890133fa7","entry":"compute_pareto_frontier","repo":"J-Rosser-UK/AgentBreeder","repo_kind":"official","path":"src/adas_eval.py","file_url":"https://github.com/J-Rosser-UK/AgentBreeder/blob/HEAD/src/adas_eval.py","link_basis":"harvester_set","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":"f8bdf17890133fa7"}},{"code_sha256_prefix":"8cd348e3c312a87b","entry":"extract_async_forward_as_string","repo":"J-Rosser-UK/AgentBreeder","repo_kind":"official","path":"src/adas_eval.py","file_url":"https://github.com/J-Rosser-UK/AgentBreeder/blob/HEAD/src/adas_eval.py","link_basis":"harvester_set","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":"8cd348e3c312a87b"}},{"code_sha256_prefix":"1d4a9efbfddea497","entry":"count_tokens","repo":"J-Rosser-UK/AgentBreeder","repo_kind":"official","path":"src/api/anthropic_api.py","file_url":"https://github.com/J-Rosser-UK/AgentBreeder/blob/HEAD/src/api/anthropic_api.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":"1d4a9efbfddea497"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}