{"paper":{"title":"Outcome-Mass Alignment: Controlling Length-Dependent Trajectory Influence in Multi-Turn LLM Agent Reinforcement Learning","arxiv_id":"2611.16816","primary_cat":"cs.CL","submitted_at":"2026-11-12T13:52:12+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":0.96,"id":"s0","own":false,"role":"background","text":"To keep long interaction contexts tractable, stepwise group-relative training for multi-turn LLM agents materializes each source trajectory into multiple response rows and valid tokens.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.26,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.05,"role_probabilities":{"gap":0.0,"method":0.04,"result":0.0,"background":0.96,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.57,"id":"s1","own":false,"role":"background","text":"This expansion creates an outer-accounting problem distinct from within-trajectory credit assignment: row-based outcome statistics can count one episode outcome multiple times, while token-level reduction can assign unequal aggregate objective mass across trajectories.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.35,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.15,"role_probabilities":{"gap":0.42,"method":0.0,"result":0.0,"background":0.57,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.01}},{"i":2,"p":0.51,"id":"s2","own":true,"role":"background","text":"We term these effects outcome-statistics multiplicity and objective-mass imbalance.","hedge":{"p":0.93,"label":"asserted"},"own_p":0.56,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.12,"method":0.03,"result":0.0,"background":0.51,"limitation":0.0,"future_work":0.0,"contribution":0.3,"interpretation":0.04}},{"i":3,"p":0.98,"id":"s3","own":true,"role":"contribution","text":"We introduce Outcome-Mass Alignment (OMA), a two-component method that makes both accounting choices explicit.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.89,"main_claim_p":0.07,"quantitative":false,"quantitative_p":0.11,"role_probabilities":{"gap":0.0,"method":0.02,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.98,"interpretation":0.0}},{"i":4,"p":0.87,"id":"s4","own":true,"role":"method","text":"Outcome Alignment computes outcome statistics over unique source trajectories, while Mass Alignment controls aggregate objective mass after within-trajectory aggregation of optimizer-native token objectives.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.84,"main_claim_p":0.1,"quantitative":false,"quantitative_p":0.05,"role_probabilities":{"gap":0.0,"method":0.87,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.13,"interpretation":0.0}},{"i":5,"p":0.68,"id":"s5","own":true,"role":"method","text":"Neither redesigns the base optimizer's credit assignment, importance ratios, or clipping.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.69,"main_claim_p":0.05,"quantitative":false,"quantitative_p":0.05,"role_probabilities":{"gap":0.0,"method":0.68,"result":0.0,"background":0.0,"limitation":0.19,"future_work":0.0,"contribution":0.13,"interpretation":0.0}},{"i":6,"p":1.0,"id":"s6","own":true,"role":"result","text":"Across matched ALFWorld GRPO comparisons at two model scales, where failed trajectories typically expand much more than successful ones, OMA improves matched-horizon endpoint success by 7.0-9.4 percentage points.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.96,"main_claim_p":0.33,"quantitative":true,"quantitative_p":0.99,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":7,"p":1.0,"id":"s7","own":true,"role":"result","text":"Ordinary WebShop exhibits weak or reversed expansion-outcome coupling and correspondingly mixed effects, whereas in a success-oriented WebShop setting, WebShop-SOL, OMA improves success nAUC from 0.399 to 0.511.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.94,"main_claim_p":0.12,"quantitative":true,"quantitative_p":0.99,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":8,"p":0.97,"id":"s8","own":true,"role":"result","text":"Component ablations and GiGPO and GSPO comparisons further support component separability and cross-optimizer compatibility.","hedge":{"p":0.95,"label":"asserted"},"own_p":0.93,"main_claim_p":0.05,"quantitative":false,"quantitative_p":0.34,"role_probabilities":{"gap":0.0,"method":0.01,"result":0.97,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.02,"interpretation":0.0}},{"i":9,"p":1.0,"id":"s9","own":true,"role":"interpretation","text":"These results show that trajectory expansion always defines an implicit weighting rule, but its practical effect depends on how expansion couples to outcomes and optimizer-native updates.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.81,"main_claim_p":0.74,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":1.0}}]},"compiled":{"main_claim":{"p":0.97,"text":"These results show that trajectory expansion always defines an implicit weighting rule, but its practical effect depends on how expansion couples to outcomes and optimizer-native updates.","location":"abstract s9","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":7649,"cost_usd_reported":0.0},"compiled_at":"2026-11-13","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.07,"states_limitation":0.81,"human_participants":0.04,"controlled_comparison":0.94,"randomized_assignment":0.03,"states_quantitative_result":0.99,"compares_with_prior_results":0.95,"mentions_code_or_data_availability":0.03},"evidential_basis":{"p":0.84,"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":0.84,"simulation_or_computation":0.16}},"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":0.99,"text":"We introduce Outcome-Mass Alignment (OMA), a two-component method that makes both accounting choices explicit.","location":"abstract s3"}},"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":10,"sentences_in_abstract":10},"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"}}}