{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:CRFJBUMYNECYNWRKXUZTF3W3A6","short_pith_number":"pith:CRFJBUMY","canonical_record":{"source":{"id":"2511.08113","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-11-11T11:11:31Z","cross_cats_sorted":[],"title_canon_sha256":"d3b917430e868284343e5818a58c7d84432e52e0952fd0d688d311d8450d4f6f","abstract_canon_sha256":"f5464ada0629e5cca318b6ead4cc30b9c0076bb96008fa696dc2f893b18bc77d"},"schema_version":"1.0"},"canonical_sha256":"144a90d198690586da2abd3332eedb07ae3b5734891cd0a393441b36c71f815f","source":{"kind":"arxiv","id":"2511.08113","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2511.08113","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"arxiv_version","alias_value":"2511.08113v3","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.08113","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"pith_short_12","alias_value":"CRFJBUMYNECY","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"pith_short_16","alias_value":"CRFJBUMYNECYNWRK","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"pith_short_8","alias_value":"CRFJBUMY","created_at":"2026-06-11T01:10:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:CRFJBUMYNECYNWRKXUZTF3W3A6","target":"record","payload":{"canonical_record":{"source":{"id":"2511.08113","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-11-11T11:11:31Z","cross_cats_sorted":[],"title_canon_sha256":"d3b917430e868284343e5818a58c7d84432e52e0952fd0d688d311d8450d4f6f","abstract_canon_sha256":"f5464ada0629e5cca318b6ead4cc30b9c0076bb96008fa696dc2f893b18bc77d"},"schema_version":"1.0"},"canonical_sha256":"144a90d198690586da2abd3332eedb07ae3b5734891cd0a393441b36c71f815f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-11T01:10:29.596568Z","signature_b64":"kRTIK5HwOXnR0xfc24JmvXbkdxN5jHunvXEL3TXEenP+h1qipUvcqmOBBQbOEHmrWF9YDqem2/5Kfrpjssi+Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"144a90d198690586da2abd3332eedb07ae3b5734891cd0a393441b36c71f815f","last_reissued_at":"2026-06-11T01:10:29.595453Z","signature_status":"signed_v1","first_computed_at":"2026-06-11T01:10:29.595453Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2511.08113","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-11T01:10:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"l60rJaL5l0z16qungZKBExpIsxexVY/6TS+5T3Jg8Kv34CxrX42TNg73WkorKRgS+Uxxw+zDCCZldhQLsHWhBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-27T21:21:06.394491Z"},"content_sha256":"59db4f64a2dc20f9d23b9523abb0a97eae0b7a3f048e166a7ddbe4e60628df2d","schema_version":"1.0","event_id":"sha256:59db4f64a2dc20f9d23b9523abb0a97eae0b7a3f048e166a7ddbe4e60628df2d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:CRFJBUMYNECYNWRKXUZTF3W3A6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Multimodal LLMs Do Not Compose Skills Optimally Across Modalities","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aitor Ormazabal, Gorka Azkune, Paula Ontalvilla","submitted_at":"2025-11-11T11:11:31Z","abstract_excerpt":"Skill composition is the ability to combine previously learned skills to solve new tasks. As neural networks acquire increasingly complex skills during their pretraining, it is not clear how successfully they can compose them. In this paper, we focus on Multimodal Large Language Models (MLLM), and study their ability to compose skills across modalities. To this end, we design three evaluation tasks which can be solved sequentially composing two modality-dependent skills, and evaluate several open MLLMs under two main settings: i) prompting the model to directly solve the task, and ii) using a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.08113","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2511.08113/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-11T01:10:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jMydoyq6R/+Z5YIfp5Fgbn+8hUUeMLxeuZFyb8d2nK7XVYFwrB3WTPOuR22mDu1madtGTHpHKf857TJl79miAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-27T21:21:06.394857Z"},"content_sha256":"da0652838d39de7d3c701b1b065e674f21b7dc63063229364732081dc43f8b9b","schema_version":"1.0","event_id":"sha256:da0652838d39de7d3c701b1b065e674f21b7dc63063229364732081dc43f8b9b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/CRFJBUMYNECYNWRKXUZTF3W3A6/bundle.json","state_url":"https://pith.science/pith/CRFJBUMYNECYNWRKXUZTF3W3A6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/CRFJBUMYNECYNWRKXUZTF3W3A6/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-27T21:21:06Z","links":{"resolver":"https://pith.science/pith/CRFJBUMYNECYNWRKXUZTF3W3A6","bundle":"https://pith.science/pith/CRFJBUMYNECYNWRKXUZTF3W3A6/bundle.json","state":"https://pith.science/pith/CRFJBUMYNECYNWRKXUZTF3W3A6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/CRFJBUMYNECYNWRKXUZTF3W3A6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:CRFJBUMYNECYNWRKXUZTF3W3A6","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"f5464ada0629e5cca318b6ead4cc30b9c0076bb96008fa696dc2f893b18bc77d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-11-11T11:11:31Z","title_canon_sha256":"d3b917430e868284343e5818a58c7d84432e52e0952fd0d688d311d8450d4f6f"},"schema_version":"1.0","source":{"id":"2511.08113","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2511.08113","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"arxiv_version","alias_value":"2511.08113v3","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.08113","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"pith_short_12","alias_value":"CRFJBUMYNECY","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"pith_short_16","alias_value":"CRFJBUMYNECYNWRK","created_at":"2026-06-11T01:10:29Z"},{"alias_kind":"pith_short_8","alias_value":"CRFJBUMY","created_at":"2026-06-11T01:10:29Z"}],"graph_snapshots":[{"event_id":"sha256:da0652838d39de7d3c701b1b065e674f21b7dc63063229364732081dc43f8b9b","target":"graph","created_at":"2026-06-11T01:10:29Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2511.08113/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Skill composition is the ability to combine previously learned skills to solve new tasks. As neural networks acquire increasingly complex skills during their pretraining, it is not clear how successfully they can compose them. In this paper, we focus on Multimodal Large Language Models (MLLM), and study their ability to compose skills across modalities. To this end, we design three evaluation tasks which can be solved sequentially composing two modality-dependent skills, and evaluate several open MLLMs under two main settings: i) prompting the model to directly solve the task, and ii) using a ","authors_text":"Aitor Ormazabal, Gorka Azkune, Paula Ontalvilla","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-11-11T11:11:31Z","title":"Multimodal LLMs Do Not Compose Skills Optimally Across Modalities"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.08113","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:59db4f64a2dc20f9d23b9523abb0a97eae0b7a3f048e166a7ddbe4e60628df2d","target":"record","created_at":"2026-06-11T01:10:29Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"f5464ada0629e5cca318b6ead4cc30b9c0076bb96008fa696dc2f893b18bc77d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-11-11T11:11:31Z","title_canon_sha256":"d3b917430e868284343e5818a58c7d84432e52e0952fd0d688d311d8450d4f6f"},"schema_version":"1.0","source":{"id":"2511.08113","kind":"arxiv","version":3}},"canonical_sha256":"144a90d198690586da2abd3332eedb07ae3b5734891cd0a393441b36c71f815f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"144a90d198690586da2abd3332eedb07ae3b5734891cd0a393441b36c71f815f","first_computed_at":"2026-06-11T01:10:29.595453Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-11T01:10:29.595453Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"kRTIK5HwOXnR0xfc24JmvXbkdxN5jHunvXEL3TXEenP+h1qipUvcqmOBBQbOEHmrWF9YDqem2/5Kfrpjssi+Bg==","signature_status":"signed_v1","signed_at":"2026-06-11T01:10:29.596568Z","signed_message":"canonical_sha256_bytes"},"source_id":"2511.08113","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:59db4f64a2dc20f9d23b9523abb0a97eae0b7a3f048e166a7ddbe4e60628df2d","sha256:da0652838d39de7d3c701b1b065e674f21b7dc63063229364732081dc43f8b9b"],"state_sha256":"e9b0ef246fe977b2a48e63d26325eff7675711c79f353d73f832f5b43f89e5b7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ld7AVfw9QSrhnb7u5TR3mpzmACM2mF80/hZZ4gIQ4eYaFqFsVaHVBrD4s235j6rGn05JrcUXbgAAAsJKpcKnDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-27T21:21:06.396768Z","bundle_sha256":"2d13941889b45bc50dd8ead598d3c5defbfb9b035f6ed5d87496e041c43aa382"}}