{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/arxiv-2605-11301","title":"LatentRouter: Can We Choose the Right Multimodal Model Before Seeing Its Answer?","arxiv_id":"2605.11301","date":"2026-05-11","proceeding":null,"authors":["Xueqi Cheng","Yushun Dong"],"abstract":"Multimodal large language models (MLLMs) have heterogeneous strengths across OCR, chart understanding, spatial reasoning, visual question answering, cost, and latency. Effective MLLM routing therefore requires more than estimating query difficulty: a router must match the multimodal requirements of the current image-question input with the capabilities of each candidate model. We propose LatentRouter, a router that formulates MLLM routing as counterfactual multimodal utility prediction. Given an image-question query, LatentRouter extracts learned multimodal routing capsules, represents each candidate MLLM with a model capability token, and performs latent communication between these states to estimate how each model would perform if selected. A distributional outcome head predicts model-specific counterfactual quality, while a bounded capsule correction refines close decisions without allowing residual signals to dominate the prediction. The resulting utility-based policy supports performance-oriented and performance-cost routing, and handles changing candidate pools through shared per-model scoring with availability masking. Experiments on MMR-Bench and VL-RouterBench show that LatentRouter outperforms fixed-model, feature-level, and learned-router baselines. Additional analyses show that the gains are strongest on multimodal task groups where model choice depends on visual, layout-sensitive, or reasoning-oriented requirements, and that latent communication is the main contributor to the improvement. The code is available at: https://github.com/LabRAI/LatentRouter.","url_abs":"https://arxiv.org/abs/2605.11301","url_pdf":"https://arxiv.org/pdf/2605.11301","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2605.11301","atlas_url":null,"mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2605.11301"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/LabRAI/LatentRouter","reach":null}],"summary":{"ran":6,"ran_draft_wrong":1,"unverified":1},"by_repo_kind":{"found_in_text":{"samples":8,"ran":7,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"d960f42aad3c72c7","entry":"discover_model_groups","repo":"LabRAI/LatentRouter","repo_kind":"found_in_text","path":"src/latentrouter/benchmarks/mmr.py","file_url":"https://github.com/LabRAI/LatentRouter/blob/HEAD/src/latentrouter/benchmarks/mmr.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d960f42aad3c72c7"}},{"code_sha256_prefix":"25ba597a53fa8704","entry":"ensure_dir","repo":"LabRAI/LatentRouter","repo_kind":"found_in_text","path":"src/latentrouter/io.py","file_url":"https://github.com/LabRAI/LatentRouter/blob/HEAD/src/latentrouter/io.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"25ba597a53fa8704"}},{"code_sha256_prefix":"f6fd3c5a4e15c317","entry":"load_run_config","repo":"LabRAI/LatentRouter","repo_kind":"found_in_text","path":"src/latentrouter/config.py","file_url":"https://github.com/LabRAI/LatentRouter/blob/HEAD/src/latentrouter/config.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f6fd3c5a4e15c317"}},{"code_sha256_prefix":"6dd9cb3d46e60860","entry":"parse_dataset_fields","repo":"LabRAI/LatentRouter","repo_kind":"found_in_text","path":"src/latentrouter/benchmarks/mmr.py","file_url":"https://github.com/LabRAI/LatentRouter/blob/HEAD/src/latentrouter/benchmarks/mmr.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6dd9cb3d46e60860"}},{"code_sha256_prefix":"2fae12bd97a83615","entry":"rank_score","repo":"LabRAI/LatentRouter","repo_kind":"found_in_text","path":"src/latentrouter/benchmarks/vl_routerbench.py","file_url":"https://github.com/LabRAI/LatentRouter/blob/HEAD/src/latentrouter/benchmarks/vl_routerbench.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2fae12bd97a83615"}},{"code_sha256_prefix":"578cfc781460c39a","entry":"read_json","repo":"LabRAI/LatentRouter","repo_kind":"found_in_text","path":"src/latentrouter/io.py","file_url":"https://github.com/LabRAI/LatentRouter/blob/HEAD/src/latentrouter/io.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"578cfc781460c39a"}},{"code_sha256_prefix":"058d208375639d01","entry":"stable_hash","repo":"LabRAI/LatentRouter","repo_kind":"found_in_text","path":"src/latentrouter/io.py","file_url":"https://github.com/LabRAI/LatentRouter/blob/HEAD/src/latentrouter/io.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"058d208375639d01"}},{"code_sha256_prefix":"0af18f548ef77495","entry":"load_frame","repo":"LabRAI/LatentRouter","repo_kind":"found_in_text","path":"src/latentrouter/benchmarks/mmr.py","file_url":"https://github.com/LabRAI/LatentRouter/blob/HEAD/src/latentrouter/benchmarks/mmr.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0af18f548ef77495"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.CL","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}