{"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/arxiv-2603-19453","title":"Beyond Scalar Rewards: Dense Feedback for LLM Policy Synthesis in Sequential Social Dilemmas","arxiv_id":"2603.19453","date":"2026-03-19","proceeding":null,"authors":["Víctor Gallego"],"abstract":"We propose an LLM harness that generates code-based policy functions for multi-agent environments, evaluates them with self-play, and refines them using feedback from previous iterations. Following the recent line of work in feedback engineering (the design of which information signals are shown to the LLM during refinement), we compare sparse feedback (scalar reward only) with dense feedback (reward plus social metrics: efficiency, equality, sustainability, peace). In two Sequential Social Dilemmas (Gathering and Cleanup) and with two frontier LLMs (Claude Sonnet 4.6, Gemini 3.1 Pro), dense feedback improves over or matches sparse feedback on all metrics. We explain this asymmetry via feedback aliasing: when the scalar reward maps distinct failure modes into the same value (e.g., under- vs. over-cleaning), social metrics disambiguate and allow the LLM to diagnose which direction of improvement to take. We conclude that social metrics act as a coordination signal, leading to strategies such as Voronoi territory partitioning and adaptive cleaner schedules. Code at https://github.com/vicgalle/llm-policies-social-dilemmas.","url_abs":"https://arxiv.org/abs/2603.19453","url_pdf":"https://arxiv.org/pdf/2603.19453","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2603.19453","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2603.19453"}},"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":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/vicgalle/llm-policies-social-dilemmas","reach":null}],"summary":{"ran":5,"unverified":2},"by_repo_kind":{"found_in_text":{"samples":7,"ran":5,"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":"71812e35b60a8a80","entry":"extract_policy_code","repo":"vicgalle/llm-policies-social-dilemmas","repo_kind":"found_in_text","path":"llm_self_play.py","file_url":"https://github.com/vicgalle/llm-policies-social-dilemmas/blob/HEAD/llm_self_play.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"71812e35b60a8a80"}},{"code_sha256_prefix":"4e913c04af0e4c4d","entry":"get_opponents","repo":"vicgalle/llm-policies-social-dilemmas","repo_kind":"found_in_text","path":"llm_self_play.py","file_url":"https://github.com/vicgalle/llm-policies-social-dilemmas/blob/HEAD/llm_self_play.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":"4e913c04af0e4c4d"}},{"code_sha256_prefix":"ec5f25200b4fe84a","entry":"parse_cleanup_map","repo":"vicgalle/llm-policies-social-dilemmas","repo_kind":"found_in_text","path":"cleanup_env.py","file_url":"https://github.com/vicgalle/llm-policies-social-dilemmas/blob/HEAD/cleanup_env.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ec5f25200b4fe84a"}},{"code_sha256_prefix":"0ee896b5cd691fee","entry":"parse_map","repo":"vicgalle/llm-policies-social-dilemmas","repo_kind":"found_in_text","path":"gathering_env.py","file_url":"https://github.com/vicgalle/llm-policies-social-dilemmas/blob/HEAD/gathering_env.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0ee896b5cd691fee"}},{"code_sha256_prefix":"32fe42a6f676d009","entry":"validate_code_safety","repo":"vicgalle/llm-policies-social-dilemmas","repo_kind":"found_in_text","path":"llm_self_play.py","file_url":"https://github.com/vicgalle/llm-policies-social-dilemmas/blob/HEAD/llm_self_play.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":"32fe42a6f676d009"}},{"code_sha256_prefix":"1fda83977326a03b","entry":"make_gathering","repo":"vicgalle/llm-policies-social-dilemmas","repo_kind":"found_in_text","path":"gathering_env.py","file_url":"https://github.com/vicgalle/llm-policies-social-dilemmas/blob/HEAD/gathering_env.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":"1fda83977326a03b"}},{"code_sha256_prefix":"7b978bb99191a524","entry":"make_gathering_large","repo":"vicgalle/llm-policies-social-dilemmas","repo_kind":"found_in_text","path":"gathering_env.py","file_url":"https://github.com/vicgalle/llm-policies-social-dilemmas/blob/HEAD/gathering_env.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":"7b978bb99191a524"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.CL","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}