{"paper":{"title":"Harness Self-Evolution for Long-Horizon Terminal Agents","arxiv_id":"2610.27435","primary_cat":"cs.AI","submitted_at":"2026-10-20T00:36:49+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":0.99,"id":"s0","own":false,"role":"background","text":"Long-horizon terminal agents may solve a task in one run yet fail across repeated runs.","hedge":{"p":1.0,"label":"hedged_possibility"},"own_p":0.19,"main_claim_p":0.04,"quantitative":false,"quantitative_p":0.2,"role_probabilities":{"gap":0.01,"method":0.0,"result":0.0,"background":0.99,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.98,"id":"s1","own":false,"role":"background","text":"Stochastic errors can propagate through actions and context, creating a gap between capability and reliable execution.","hedge":{"p":0.78,"label":"asserted"},"own_p":0.1,"main_claim_p":0.04,"quantitative":false,"quantitative_p":0.15,"role_probabilities":{"gap":0.02,"method":0.0,"result":0.0,"background":0.98,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":0.96,"id":"s2","own":true,"role":"contribution","text":"We present RiseHarness, a framework for iteratively improving agent harnesses with model weights fixed.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.89,"main_claim_p":0.09,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.04,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.96,"interpretation":0.0}},{"i":3,"p":0.94,"id":"s3","own":true,"role":"method","text":"It focuses on flipping tasks with both successful and failed runs under the same configuration, using successful runs as references for failure analysis.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.74,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.36,"role_probabilities":{"gap":0.0,"method":0.94,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.06,"interpretation":0.0}},{"i":4,"p":0.99,"id":"s4","own":true,"role":"method","text":"RiseHarness combines agent trajectories with internal harness traces from a runtime monitor to examine how model actions interact with command execution, feedback, context management, and completion decisions.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.85,"main_claim_p":0.04,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.99,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.01,"interpretation":0.0}},{"i":5,"p":0.93,"id":"s5","own":true,"role":"method","text":"A Harness Handbook links this evidence to harness behaviors and source code, helping identify recurring problems in mechanisms shared across tasks.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.81,"main_claim_p":0.04,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.93,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.07,"interpretation":0.0}},{"i":6,"p":0.99,"id":"s6","own":true,"role":"method","text":"RiseHarness then searches other harness frameworks for relevant implementations and adapts them as design references to the current model and execution protocol.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.84,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.99,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.01,"interpretation":0.0}},{"i":7,"p":1.0,"id":"s7","own":true,"role":"method","text":"Candidate changes undergo behavioral checks and repeated task evaluation; versions with measured gains seed the next iteration.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.89,"main_claim_p":0.08,"quantitative":false,"quantitative_p":0.11,"role_probabilities":{"gap":0.0,"method":1.0,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":8,"p":0.99,"id":"s8","own":true,"role":"method","text":"This loop integrates suitable mechanisms from different frameworks into a single harness and uses new execution evidence to guide further changes.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.82,"main_claim_p":0.1,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.99,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.01,"interpretation":0.0}},{"i":9,"p":0.52,"id":"s9","own":true,"role":"method","text":"We evaluate task success and reliability across models, ablate harness execution evidence and external research, and test transfer to unseen benchmarks with the same model.","hedge":{"p":0.96,"label":"asserted"},"own_p":0.95,"main_claim_p":0.12,"quantitative":false,"quantitative_p":0.14,"role_probabilities":{"gap":0.0,"method":0.52,"result":0.03,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.45,"interpretation":0.0}},{"i":10,"p":0.75,"id":"s10","own":true,"role":"contribution","text":"RiseHarness treats a specific model as the unit of harness adaptation, studying how changes to shared interaction mechanisms can make existing capabilities more reliable across tasks.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.7,"main_claim_p":0.28,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":0.0,"method":0.18,"result":0.0,"background":0.0,"limitation":0.01,"future_work":0.0,"contribution":0.75,"interpretation":0.06}}]},"compiled":{"main_claim":{"p":0.85,"text":"RiseHarness treats a specific model as the unit of harness adaptation, studying how changes to shared interaction mechanisms can make existing capabilities more reliable across tasks.","location":"abstract s10","hedge_strength":{"p":0.98,"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":8078,"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":0.05,"states_limitation":0.32,"human_participants":0.03,"controlled_comparison":0.29,"randomized_assignment":0.02,"states_quantitative_result":0.04,"compares_with_prior_results":0.48,"mentions_code_or_data_availability":0.04},"evidential_basis":{"p":0.99,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.0,"analysis_of_existing_data":0.0,"observation_or_experiment":0.99,"simulation_or_computation":0.01}},"contribution_type":{"p":1.0,"value":"method or tool","probabilities":{"review":0.0,"method_or_tool":1.0,"empirical_study":0.0,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":1.0,"text":"We present RiseHarness, a framework for iteratively improving agent harnesses with model weights fixed.","location":"abstract s2"}},"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":11,"sentences_in_abstract":11},"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"}}}