{"paper":{"title":"SkillLift: Learning Dense Rubrics from Sparse Oracles for Efficient Skill Evolution","arxiv_id":"2609.15396","primary_cat":"cs.AI","submitted_at":"2026-09-14T11:30:23+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":"LLM-based agents increasingly rely on persistent skills, i.e., reusable procedural prompts, to adapt without weight updates.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.05,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.03,"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":0.69,"id":"s1","own":false,"role":"gap","text":"Existing skill self-evolution methods directly revise skill text based on execution feedback, but each oracle evaluation requires a full agent rollout, creating a supervision bottleneck that confines search to failure-patching updates.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.05,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.1,"role_probabilities":{"gap":0.69,"method":0.0,"result":0.0,"background":0.3,"limitation":0.01,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":0.64,"id":"s2","own":false,"role":"contribution","text":"Our key insight is that ranking is a smoother supervision target than absolute outcome regression: identifying which skill is better requires fewer oracle evaluations than predicting exact scores.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.41,"main_claim_p":0.15,"quantitative":true,"quantitative_p":0.58,"role_probabilities":{"gap":0.03,"method":0.15,"result":0.0,"background":0.06,"limitation":0.0,"future_work":0.0,"contribution":0.64,"interpretation":0.12}},{"i":3,"p":0.84,"id":"s3","own":true,"role":"contribution","text":"Building on this insight, we propose SkillLift, which decouples skill search from oracle cost by learning an oracle-aligned rubric as a structured evaluation space.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.89,"main_claim_p":0.15,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.16,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.84,"interpretation":0.0}},{"i":4,"p":0.99,"id":"s4","own":true,"role":"method","text":"We formalize this as a bilevel optimization problem solved via alternating optimization: an inner loop uses the frozen rubric as a cheap surrogate to guide skill revision at no oracle cost, while an outer loop invokes a small number of oracle rollouts to re-align the rubric via rank correlation, amortizing oracle cost and stabilizing text-space updates.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.92,"main_claim_p":0.11,"quantitative":false,"quantitative_p":0.1,"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":1.0,"id":"s5","own":true,"role":"result","text":"Experiments on complex agent task benchmarks show that our method outperforms existing auto-skill methods with 40--70\\% less token cost compared to frontier evolving methods.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.96,"main_claim_p":0.59,"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":6,"p":0.63,"id":"s6","own":false,"role":"contribution","text":"Codes are available at https://github.com/WalteR-MittY-pro/SkillLift.","hedge":{"p":0.78,"label":"asserted"},"own_p":0.42,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.06,"result":0.06,"background":0.02,"limitation":0.0,"future_work":0.23,"contribution":0.63,"interpretation":0.0}}]},"compiled":{"main_claim":{"p":0.99,"text":"Experiments on complex agent task benchmarks show that our method outperforms existing auto-skill methods with 40--70\\% less token cost compared to frontier evolving methods.","location":"abstract s5","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":5710,"cost_usd_reported":0.0},"compiled_at":"2026-09-23","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.06,"states_limitation":0.11,"human_participants":0.04,"controlled_comparison":0.74,"randomized_assignment":0.01,"states_quantitative_result":0.98,"compares_with_prior_results":0.97,"mentions_code_or_data_availability":0.99},"evidential_basis":{"p":0.97,"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":0.97,"simulation_or_computation":0.03}},"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":0.96,"text":"Building on this insight, we propose SkillLift, which decouples skill search from oracle cost by learning an oracle-aligned rubric as a structured evaluation space.","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":7,"sentences_in_abstract":7},"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"}}}