{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:WXUCKGPRV4NPP3HBWKXERJUHNO","short_pith_number":"pith:WXUCKGPR","canonical_record":{"source":{"id":"2602.00462","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-31T02:33:07Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c35a17c4fa2fe3ca4aa1a4f26c064153314c59dc7721dbbf9fc2aa72d97e35cc","abstract_canon_sha256":"471996a322e2d246bd9b70d2472aea331381cec0e17da350b1e63035bafe81b5"},"schema_version":"1.0"},"canonical_sha256":"b5e82519f1af1af7ece1b2ae48a6876bb043147bbe19ab64fef6adf4bd216b64","source":{"kind":"arxiv","id":"2602.00462","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.00462","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"arxiv_version","alias_value":"2602.00462v4","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.00462","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"pith_short_12","alias_value":"WXUCKGPRV4NP","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"pith_short_16","alias_value":"WXUCKGPRV4NPP3HB","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"pith_short_8","alias_value":"WXUCKGPR","created_at":"2026-06-12T01:08:22Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:WXUCKGPRV4NPP3HBWKXERJUHNO","target":"record","payload":{"canonical_record":{"source":{"id":"2602.00462","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-31T02:33:07Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c35a17c4fa2fe3ca4aa1a4f26c064153314c59dc7721dbbf9fc2aa72d97e35cc","abstract_canon_sha256":"471996a322e2d246bd9b70d2472aea331381cec0e17da350b1e63035bafe81b5"},"schema_version":"1.0"},"canonical_sha256":"b5e82519f1af1af7ece1b2ae48a6876bb043147bbe19ab64fef6adf4bd216b64","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-12T01:08:22.053386Z","signature_b64":"Y/cbSXdLqrx95vMPm64GAjL0h58ikADBKFppQdmZx/6cUKJuTCn0vJojwQABlXvZdKJFx7jvLjnaagpomJ4rCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b5e82519f1af1af7ece1b2ae48a6876bb043147bbe19ab64fef6adf4bd216b64","last_reissued_at":"2026-06-12T01:08:22.052146Z","signature_status":"signed_v1","first_computed_at":"2026-06-12T01:08:22.052146Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.00462","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-12T01:08:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KZsERTqPgTq992yNM7KclghfH+YcD1gUxomdJfDHcKryeEyR4t8otTc3oQwh0/mEGin5qpeHi11is5t1TIh8Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-02T21:46:15.379880Z"},"content_sha256":"dc787e24ef6427e3b8b03848992d68b09c444146f5feb2734b320e1ed8033640","schema_version":"1.0","event_id":"sha256:dc787e24ef6427e3b8b03848992d68b09c444146f5feb2734b320e1ed8033640"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:WXUCKGPRV4NPP3HBWKXERJUHNO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LatentLens: Revealing Highly Interpretable Visual Tokens in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Benno Krojer, Desmond Elliott, Marius Mosbach, Oscar Ma\\~nas, Shravan Nayak, Siva Reddy, Vaibhav Adlakha","submitted_at":"2026-01-31T02:33:07Z","abstract_excerpt":"Transforming a large language model (LLM) into a vision-language model (VLM) can be achieved by mapping the visual tokens from a vision encoder into the embedding space of an LLM. Intriguingly, this mapping can be as simple as a shallow MLP transformation. To understand why LLMs can so readily process visual tokens, we need interpretability methods that reveal what is encoded in the visual token representations at every layer of LLM processing. In this work, we introduce LatentLens, a novel approach for mapping latent representations to descriptions in natural language. LatentLens encodes a la"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.00462","kind":"arxiv","version":4},"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/2602.00462/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":"2026-06-12T01:08:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DUdftH3N8NkYp5MjozK6+G4vHUs3/33wNq6/mTQMVFTHzXzwsuCiWuzkGlEzUSC6ar0O4PjVDQzHG+s9Ja8+Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-02T21:46:15.380261Z"},"content_sha256":"6bcd57c84099d572e36a69d2f4c19851a01d9e619c553e57ecd3b29e92ac700c","schema_version":"1.0","event_id":"sha256:6bcd57c84099d572e36a69d2f4c19851a01d9e619c553e57ecd3b29e92ac700c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WXUCKGPRV4NPP3HBWKXERJUHNO/bundle.json","state_url":"https://pith.science/pith/WXUCKGPRV4NPP3HBWKXERJUHNO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WXUCKGPRV4NPP3HBWKXERJUHNO/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":"2026-07-02T21:46:15Z","links":{"resolver":"https://pith.science/pith/WXUCKGPRV4NPP3HBWKXERJUHNO","bundle":"https://pith.science/pith/WXUCKGPRV4NPP3HBWKXERJUHNO/bundle.json","state":"https://pith.science/pith/WXUCKGPRV4NPP3HBWKXERJUHNO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WXUCKGPRV4NPP3HBWKXERJUHNO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:WXUCKGPRV4NPP3HBWKXERJUHNO","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":"471996a322e2d246bd9b70d2472aea331381cec0e17da350b1e63035bafe81b5","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-31T02:33:07Z","title_canon_sha256":"c35a17c4fa2fe3ca4aa1a4f26c064153314c59dc7721dbbf9fc2aa72d97e35cc"},"schema_version":"1.0","source":{"id":"2602.00462","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.00462","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"arxiv_version","alias_value":"2602.00462v4","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.00462","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"pith_short_12","alias_value":"WXUCKGPRV4NP","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"pith_short_16","alias_value":"WXUCKGPRV4NPP3HB","created_at":"2026-06-12T01:08:22Z"},{"alias_kind":"pith_short_8","alias_value":"WXUCKGPR","created_at":"2026-06-12T01:08:22Z"}],"graph_snapshots":[{"event_id":"sha256:6bcd57c84099d572e36a69d2f4c19851a01d9e619c553e57ecd3b29e92ac700c","target":"graph","created_at":"2026-06-12T01:08:22Z","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/2602.00462/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Transforming a large language model (LLM) into a vision-language model (VLM) can be achieved by mapping the visual tokens from a vision encoder into the embedding space of an LLM. Intriguingly, this mapping can be as simple as a shallow MLP transformation. To understand why LLMs can so readily process visual tokens, we need interpretability methods that reveal what is encoded in the visual token representations at every layer of LLM processing. In this work, we introduce LatentLens, a novel approach for mapping latent representations to descriptions in natural language. LatentLens encodes a la","authors_text":"Benno Krojer, Desmond Elliott, Marius Mosbach, Oscar Ma\\~nas, Shravan Nayak, Siva Reddy, Vaibhav Adlakha","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-31T02:33:07Z","title":"LatentLens: Revealing Highly Interpretable Visual Tokens in LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.00462","kind":"arxiv","version":4},"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:dc787e24ef6427e3b8b03848992d68b09c444146f5feb2734b320e1ed8033640","target":"record","created_at":"2026-06-12T01:08:22Z","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":"471996a322e2d246bd9b70d2472aea331381cec0e17da350b1e63035bafe81b5","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-31T02:33:07Z","title_canon_sha256":"c35a17c4fa2fe3ca4aa1a4f26c064153314c59dc7721dbbf9fc2aa72d97e35cc"},"schema_version":"1.0","source":{"id":"2602.00462","kind":"arxiv","version":4}},"canonical_sha256":"b5e82519f1af1af7ece1b2ae48a6876bb043147bbe19ab64fef6adf4bd216b64","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b5e82519f1af1af7ece1b2ae48a6876bb043147bbe19ab64fef6adf4bd216b64","first_computed_at":"2026-06-12T01:08:22.052146Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-12T01:08:22.052146Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Y/cbSXdLqrx95vMPm64GAjL0h58ikADBKFppQdmZx/6cUKJuTCn0vJojwQABlXvZdKJFx7jvLjnaagpomJ4rCw==","signature_status":"signed_v1","signed_at":"2026-06-12T01:08:22.053386Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.00462","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:dc787e24ef6427e3b8b03848992d68b09c444146f5feb2734b320e1ed8033640","sha256:6bcd57c84099d572e36a69d2f4c19851a01d9e619c553e57ecd3b29e92ac700c"],"state_sha256":"516148d09e9cdeae021aec5adda0ee872e69f8fce0518850b15f5c7e4c6448a4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"B64co/NkdN3G4lq/zjsXMHdOhTiwtfdqmec/r04ckKvRUBx2ORlBSQ0UqxTGP36E81EDJSXKN/CYGwc3IBpHCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-02T21:46:15.382371Z","bundle_sha256":"e218baa838cba352119ed7b4d732bd5a5152049b141d98759a2b55922cd5ba2a"}}