{"paper":{"title":"VABench: Measuring Embodied Spatial Intelligence through Visual Demonstrations, Active Perception, and Metric Control","arxiv_id":"2609.19554","primary_cat":"cs.RO","submitted_at":"2026-09-17T01:24:51+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":0.92,"id":"s0","own":false,"role":"background","text":"Spatial intelligence requires more than describing object locations.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.06,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.08,"role_probabilities":{"gap":0.08,"method":0.0,"result":0.0,"background":0.92,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.96,"id":"s1","own":false,"role":"background","text":"Under incomplete observation, models must identify and acquire missing evidence, interpret it in a common spatial frame, and act on it.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.07,"main_claim_p":0.04,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.04,"method":0.0,"result":0.0,"background":0.96,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":1.0,"id":"s2","own":true,"role":"contribution","text":"We introduce VA-Bench to evaluate the complete observe-reason-act-revise loop.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.91,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.04,"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":3,"p":0.95,"id":"s3","own":true,"role":"method","text":"General-purpose MLLMs learn procedural context from RGB-only demonstrations, actively select camera viewpoints, issue metric Cartesian commands, and revise them from execution feedback.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.65,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.06,"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":4,"p":0.99,"id":"s4","own":true,"role":"method","text":"Models receive no privileged object poses, oracle trajectories, or learned action heads.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.6,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.12,"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":0.99,"id":"s5","own":true,"role":"method","text":"A fixed model-agnostic controller executes only model-specified targets.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.54,"main_claim_p":0.02,"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":6,"p":0.81,"id":"s6","own":true,"role":"method","text":"VA-Bench contains 14 base task families (11 single-arm and three dual-arm), seven held-out geometry/layout variants, and a long-horizon five-object composition track.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.81,"main_claim_p":0.02,"quantitative":true,"quantitative_p":0.97,"role_probabilities":{"gap":0.0,"method":0.81,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.19,"interpretation":0.0}},{"i":7,"p":0.97,"id":"s7","own":true,"role":"method","text":"We evaluate 12 primary model conditions in three independent runs over the same 20 physically verified seeds per base task, reporting terminal success, nine trajectory-level behavioral diagnostics, and subtask progress.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.96,"main_claim_p":0.03,"quantitative":true,"quantitative_p":0.96,"role_probabilities":{"gap":0.0,"method":0.97,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.03,"interpretation":0.0}},{"i":8,"p":1.0,"id":"s8","own":true,"role":"result","text":"First, the best-performing model scores 100.0% on target localization and 78.9% on spatial relations in the annotated run.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.93,"main_claim_p":0.06,"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":9,"p":1.0,"id":"s9","own":true,"role":"result","text":"Its three-run macro-average task success is only 53.93+/-3.17%.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.95,"main_claim_p":0.1,"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":10,"p":0.99,"id":"s10","own":true,"role":"result","text":"Second, active camera control significantly improves task success over passive multi-view observation.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.93,"main_claim_p":0.14,"quantitative":true,"quantitative_p":0.88,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.99,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.01}},{"i":11,"p":1.0,"id":"s11","own":true,"role":"result","text":"In one matched comparison, success rises from 27.86% to 57.50%.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.94,"main_claim_p":0.06,"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":12,"p":1.0,"id":"s12","own":true,"role":"result","text":"Third, held-out geometric transfer can reduce task success by over 30 percentage points.","hedge":{"p":0.91,"label":"asserted"},"own_p":0.93,"main_claim_p":0.07,"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":13,"p":0.99,"id":"s13","own":true,"role":"result","text":"No model completes a strict long-horizon episode, despite substantial partial progress.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.93,"main_claim_p":0.19,"quantitative":false,"quantitative_p":0.21,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.99,"background":0.0,"limitation":0.01,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":14,"p":0.6,"id":"s14","own":true,"role":"interpretation","text":"VA-Bench thus tests whether general-purpose MLLMs can turn visual demonstrations and actively acquired evidence into successful embodied action.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.51,"main_claim_p":0.42,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.01,"method":0.01,"result":0.0,"background":0.01,"limitation":0.0,"future_work":0.0,"contribution":0.37,"interpretation":0.6}}]},"compiled":{"main_claim":{"p":0.49,"text":"VA-Bench thus tests whether general-purpose MLLMs can turn visual demonstrations and actively acquired evidence into successful embodied action.","location":"abstract s14","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":10604,"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":0.09,"states_limitation":0.83,"human_participants":0.05,"controlled_comparison":0.94,"randomized_assignment":0.03,"states_quantitative_result":0.99,"compares_with_prior_results":0.84,"mentions_code_or_data_availability":0.05},"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":"dataset or benchmark","probabilities":{"review":0.0,"method_or_tool":0.0,"empirical_study":0.0,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":1.0,"position_or_commentary":0.0}},"contribution_statement":{"p":1.0,"text":"We introduce VA-Bench to evaluate the complete observe-reason-act-revise loop.","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":15,"sentences_in_abstract":15},"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"}}}