{"paper":{"title":"Elicitation or Expansion? Demystifying Gains from Agent Harness Optimization","arxiv_id":"2610.27889","primary_cat":"cs.LG","submitted_at":"2026-10-20T09:11:39+00:00","version_compiled":"v1"},"notice":"Every quoted sentence is the authors' text from the arXiv listing abstract. Labels and probabilities are the outputs of classifier-1.13.0 on the question bank abstract-p1-20260923, exactly as returned. No sentence in this record was written by a language model.","schema":"pith.compiled-record/v0","status":"compiled","abstract":{"source":"arXiv listing abstract v1","sentences":[{"i":0,"p":1.0,"id":"s0","own":false,"role":"background","text":"Recent studies in harness optimization have shown substantial performance gains in language model agents without updating model weights.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.03,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.24,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":1.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":1.0,"id":"s1","own":false,"role":"background","text":"A well-designed harness more effectively leverages the underlying model's capabilities by shaping context, tool use, and execution workflows.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.06,"main_claim_p":0.04,"quantitative":false,"quantitative_p":0.05,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":1.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":0.87,"id":"s2","own":false,"role":"gap","text":"However, na\\\"ive success rates alone do not fully characterize whether optimized harness yields more reliable success on tasks already within reach (elicitation) or a broader range of solvable tasks (expansion).","hedge":{"p":0.98,"label":"asserted"},"own_p":0.09,"main_claim_p":0.07,"quantitative":false,"quantitative_p":0.13,"role_probabilities":{"gap":0.87,"method":0.0,"result":0.0,"background":0.13,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":3,"p":0.79,"id":"s3","own":false,"role":"background","text":"This distinction is crucial for recursive self-improvement and evolutionary search, as expansion may provide a broader range of successful experiences to push the capability boundaries.","hedge":{"p":0.9,"label":"hedged_possibility"},"own_p":0.05,"main_claim_p":0.07,"quantitative":false,"quantitative_p":0.14,"role_probabilities":{"gap":0.04,"method":0.0,"result":0.0,"background":0.79,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.17}},{"i":4,"p":1.0,"id":"s4","own":false,"role":"gap","text":"Despite the growing significance, this distinction remains underexplored in harness optimization.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.04,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":1.0,"method":0.0,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":5,"p":1.0,"id":"s5","own":true,"role":"contribution","text":"We conduct a systematic analysis to study the effect of gains from harness optimization.","hedge":{"p":0.87,"label":"asserted"},"own_p":0.91,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":1.0,"interpretation":0.0}},{"i":6,"p":0.95,"id":"s6","own":true,"role":"method","text":"Building on pass@k analyses in the reinforcement learning (RL) literature that compare models before and after RL training, we compare optimized harnesses with their pre-optimization counterparts.","hedge":{"p":0.93,"label":"asserted"},"own_p":0.73,"main_claim_p":0.03,"quantitative":true,"quantitative_p":0.53,"role_probabilities":{"gap":0.0,"method":0.95,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.05,"interpretation":0.0}},{"i":7,"p":0.88,"id":"s7","own":true,"role":"method","text":"Specifically, we assess changes in reachable task coverage, compare harness optimization with model scaling and increased inference-time compute, and examine how gains vary with task difficulty.","hedge":{"p":0.82,"label":"asserted"},"own_p":0.76,"main_claim_p":0.06,"quantitative":true,"quantitative_p":0.52,"role_probabilities":{"gap":0.0,"method":0.88,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.12,"interpretation":0.0}},{"i":8,"p":0.98,"id":"s8","own":true,"role":"result","text":"Across static prediction and dynamic tasks, we find that pass@1 advantages of harness over stronger interventions can reverse at larger attempt budgets and aggregate gains conceal shortfalls on hard tasks, underscoring the need for a fine-grained diagnosis.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.96,"main_claim_p":0.8,"quantitative":true,"quantitative_p":0.85,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.98,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.02}}]},"compiled":{"main_claim":{"p":1.0,"text":"Across static prediction and dynamic tasks, we find that pass@1 advantages of harness over stronger interventions can reverse at larger attempt budgets and aggregate gains conceal shortfalls on hard tasks, underscoring the need for a fine-grained diagnosis.","location":"abstract s8","hedge_strength":{"p":1.0,"label":"asserted"}},"provenance":{"route":{"via":"TypeSafe API","model":"classifier-1.13.0","endpoint":"https://api.typesafe.ai/v1/systemone","provider":"TypeSafe"},"packing":"one request per paper","requests":1,"input_tokens":6889,"cost_usd_reported":0.0},"compiled_at":"2026-10-21","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.09,"states_limitation":0.46,"human_participants":0.04,"controlled_comparison":0.92,"randomized_assignment":0.02,"states_quantitative_result":0.07,"compares_with_prior_results":0.92,"mentions_code_or_data_availability":0.02},"evidential_basis":{"p":0.89,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.0,"analysis_of_existing_data":0.05,"observation_or_experiment":0.89,"simulation_or_computation":0.06}},"contribution_type":{"p":1.0,"value":"empirical study","probabilities":{"review":0.0,"method_or_tool":0.0,"empirical_study":1.0,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":0.52,"text":"We conduct a systematic analysis to study the effect of gains from harness optimization.","location":"abstract s5"}},"coverage":{"tier":"abstract","levels_run":["P abstract"],"not_examined":["full text","figures and tables","references and citing sentences","relations to other papers"],"sources_read":["arXiv listing abstract v1"],"levels_not_run":["S sections","T body sentences","C citations","R cross-paper relations"],"sentences_examined":9,"sentences_in_abstract":9},"measured_accuracy":{"url":"https://pith.science/accuracy","measured_on":"2026-10-02","method":"two independent model families labelled a fresh random sample blind; the figure is how often the published label matches their consensus","sentence_role_agreement":0.92,"main_claim_agreement":0.908,"typed_citation_agreement":0.933,"typed_citation_agreement_95_interval":[0.917,0.945],"review_comments_accurate":0.758,"review_comments_accurate_95_interval":[0.684,0.82],"review_caught_admitted_error_found_or_related":0.448,"review_comments_judged":{"judged_by":["gpt-5.6-luna","grok-4.7"],"judged_against":"Pith's own text copy of each paper: text extracted from the PDF, the first 90,000 characters","measured_on":"2026-10-04","accurate":113,"decided":149,"sampled_reviews":200,"split_by_lane":false,"by_lane":"https://pith.science/accuracy#by-lane"}}}