{"paper":{"title":"UniSpark: Exploring Hybrid Attention for Multimodal Understanding and Generation","arxiv_id":"2611.16517","primary_cat":"cs.CV","submitted_at":"2026-11-12T09:16:10+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":true,"role":"contribution","text":"We present UniSpark-9B, a unified multimodal model that investigates hybrid attention for visual understanding and generation.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.93,"main_claim_p":0.07,"quantitative":true,"quantitative_p":0.89,"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":1,"p":0.98,"id":"s1","own":true,"role":"method","text":"Built on a Mixture-of-Transformer-Experts backbone with approximately 9B parameters in each branch, UniSpark combines full attention for explicit token interactions with GatedDeltaNet linear attention for compact context propagation.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.83,"main_claim_p":0.04,"quantitative":true,"quantitative_p":0.96,"role_probabilities":{"gap":0.0,"method":0.98,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.02,"interpretation":0.0}},{"i":2,"p":0.65,"id":"s2","own":true,"role":"method","text":"Autoregressive text prediction and visual flow matching jointly support visual question answering, text-to-image generation, instruction-guided editing, and interleaved text-image prediction.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.81,"main_claim_p":0.05,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.65,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.35,"interpretation":0.0}},{"i":3,"p":0.98,"id":"s3","own":true,"role":"method","text":"In interleaved generation, noisy visual targets are isolated from subsequent predictions, while clean visual representations provide context for later outputs.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.75,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.98,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.02,"interpretation":0.0}},{"i":4,"p":1.0,"id":"s4","own":true,"role":"method","text":"A blockwise protocol enforces this distinction through attention masks in full-attention layers and recurrent-state retention rules in linear-attention layers.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.82,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.04,"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":5,"p":1.0,"id":"s5","own":true,"role":"method","text":"The selected training sources contain 159.7M generation samples, 213.4M understanding samples, and 624.8K interleaved sequences.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.91,"main_claim_p":0.02,"quantitative":true,"quantitative_p":0.99,"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":6,"p":0.97,"id":"s6","own":true,"role":"method","text":"A four-stage curriculum jointly trains multimodal understanding and visual generation, refining answer accuracy and instruction following while extending image synthesis to 2K.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.93,"main_claim_p":0.05,"quantitative":true,"quantitative_p":0.93,"role_probabilities":{"gap":0.0,"method":0.97,"result":0.01,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.02,"interpretation":0.0}},{"i":7,"p":1.0,"id":"s7","own":true,"role":"result","text":"UniSpark scores 2470.3 on MME-S and 0.83 on GenEval.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.94,"main_claim_p":0.06,"quantitative":true,"quantitative_p":0.98,"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":0.93,"id":"s8","own":true,"role":"result","text":"Qualitative examples further illustrate its capabilities in interleaved image generation, video synthesis, and video editing.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.84,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.93,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.06,"interpretation":0.01}},{"i":9,"p":1.0,"id":"s9","own":true,"role":"interpretation","text":"Together, these results demonstrate that hybrid attention provides an effective backbone for unified multimodal understanding and generation.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.71,"main_claim_p":0.94,"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":0.0,"interpretation":1.0}}]},"compiled":{"main_claim":{"p":1.0,"text":"Together, these results demonstrate that hybrid attention provides an effective backbone for unified multimodal understanding and generation.","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":7570,"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":0.04,"states_limitation":0.05,"human_participants":0.03,"controlled_comparison":0.03,"randomized_assignment":0.01,"states_quantitative_result":0.99,"compares_with_prior_results":0.07,"mentions_code_or_data_availability":0.03},"evidential_basis":{"p":0.9,"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":0.9,"simulation_or_computation":0.1}},"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":0.84,"text":"We present UniSpark-9B, a unified multimodal model that investigates hybrid attention for visual understanding and generation.","location":"abstract s0"}},"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"}}}