{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:6ONAGNNXV7PBJCVS34VTAMMXYU","short_pith_number":"pith:6ONAGNNX","canonical_record":{"source":{"id":"2605.15183","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T17:58:27Z","cross_cats_sorted":[],"title_canon_sha256":"7bde740558cb6b4bd2e9b8c8ef41968c551bd3f797b08f522c06d4d8b1fe8847","abstract_canon_sha256":"dcce65d5f71ac5fafe019f00c52eff421f7b4aecff078765275eb1bedcc95775"},"schema_version":"1.0"},"canonical_sha256":"f39a0335b7afde148ab2df2b303197c5020b979c1c18e65cd4daded5eddefac7","source":{"kind":"arxiv","id":"2605.15183","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.15183","created_at":"2026-05-17T21:18:32Z"},{"alias_kind":"arxiv_version","alias_value":"2605.15183v1","created_at":"2026-05-17T21:18:32Z"},{"alias_kind":"pith_short_12","alias_value":"6ONAGNNXV7PB","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"6ONAGNNXV7PBJCVS","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"6ONAGNNX","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:6ONAGNNXV7PBJCVS34VTAMMXYU","target":"record","payload":{"canonical_record":{"source":{"id":"2605.15183","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T17:58:27Z","cross_cats_sorted":[],"title_canon_sha256":"7bde740558cb6b4bd2e9b8c8ef41968c551bd3f797b08f522c06d4d8b1fe8847","abstract_canon_sha256":"dcce65d5f71ac5fafe019f00c52eff421f7b4aecff078765275eb1bedcc95775"},"schema_version":"1.0"},"canonical_sha256":"f39a0335b7afde148ab2df2b303197c5020b979c1c18e65cd4daded5eddefac7","receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.2","canonical_sha256":"f39a0335b7afde148ab2df2b303197c5020b979c1c18e65cd4daded5eddefac7","last_reissued_at":"2026-05-17T21:57:18.510566Z","signature_status":"unsigned_v0","first_computed_at":"2026-05-17T21:40:25.131488Z"},"source_kind":"arxiv","source_id":"2605.15183","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T21:18:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Mx3xbjM4fRjPKIKAVzg9cxckZhTsP9R1LlFwqsm/7o9o7E5sfzCGhHTzKlnr0cx+G/eH4NRjSJr2NPY6JjAeDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-19T00:00:46.460751Z"},"content_sha256":"4a11b3d8d76acddbc7de8ee2451cce4f2120bb639cbe39da40caefdc4ba78b18","schema_version":"1.0","event_id":"sha256:4a11b3d8d76acddbc7de8ee2451cce4f2120bb639cbe39da40caefdc4ba78b18"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:6ONAGNNXV7PBJCVS34VTAMMXYU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"When Are Two Networks the Same? Tensor Similarity for Mechanistic Interpretability","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Tensor similarity is a weight-based metric that algebraically determines when two neural networks implement the same computation by ignoring irrelevant symmetries.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jacob Meyer Cohen, Laurence Wroe, Logan Riggs Smith, Melwina Albuquerque, ML Nissen Gonzalez, Thomas Dooms","submitted_at":"2026-05-14T17:58:27Z","abstract_excerpt":"Mechanistic interpretability aims to break models into meaningful parts; verifying that two such parts implement the same computation is a prerequisite. Existing similarity measures evaluate either empirical behaviour, leaving them blind to out-of-distribution mechanisms, or basis-dependent parameters, meaning they disregard weight-space symmetries. To address these issues for the class of tensor-based models, we introduce a weight-based metric, tensor similarity, that is invariant to such symmetries. This metric captures global functional equivalence and accounts for cross-layer mechanisms us"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Tensor similarity captures global functional equivalence and accounts for cross-layer mechanisms using an efficient recursive algorithm. This reduces measuring similarity and verifying faithfulness into a solved algebraic problem rather than one of empirical approximation.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the recursive algorithm correctly identifies functional equivalence for all tensor-based models without missing non-linear interactions or symmetries outside weight-space basis changes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Tensor similarity is a symmetry-invariant metric that measures functional equivalence between tensor-based networks using a recursive algorithm for cross-layer mechanisms.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Tensor similarity is a weight-based metric that algebraically determines when two neural networks implement the same computation by ignoring irrelevant symmetries.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"4890af74f90e38a0855cccf7b6724135c0954aa81b1e826dd7cd32d69f70e799"},"source":{"id":"2605.15183","kind":"arxiv","version":1},"verdict":{"id":"77df47ca-3fa3-4e3d-acf6-45a3b304390e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T03:13:32.984659Z","strongest_claim":"Tensor similarity captures global functional equivalence and accounts for cross-layer mechanisms using an efficient recursive algorithm. This reduces measuring similarity and verifying faithfulness into a solved algebraic problem rather than one of empirical approximation.","one_line_summary":"Tensor similarity is a symmetry-invariant metric that measures functional equivalence between tensor-based networks using a recursive algorithm for cross-layer mechanisms.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the recursive algorithm correctly identifies functional equivalence for all tensor-based models without missing non-linear interactions or symmetries outside weight-space basis changes.","pith_extraction_headline":"Tensor similarity is a weight-based metric that algebraically determines when two neural networks implement the same computation by ignoring irrelevant symmetries."},"references":{"count":30,"sample":[{"doi":"","year":2025,"title":"David Bau, Bolei Zhou, Aditya Khosla, Aude Oliva, and Antonio Torralba","work_id":"823a63dc-e714-4553-a714-37fe028b7fce","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Available: https://arxiv.org/abs/1704.05796","work_id":"4ecff60a-03d5-476d-b6b9-cba59c37a41c","ref_index":2,"cited_arxiv_id":"1704.05796","is_internal_anchor":true},{"doi":"","year":null,"title":"Neural networks learn statistics of increasing complexity","work_id":"3803730c-4a5c-4f3e-aafc-3ff58e1860a6","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Convolutional Rectifier Networks as Generalized Tensor Decompositions","work_id":"bc48707b-0c47-4ea8-91b0-97b76c7d0dbe","ref_index":4,"cited_arxiv_id":"1603.00162","is_internal_anchor":true},{"doi":"","year":null,"title":"On the Expressive Power of Deep Learning: A Tensor Analysis","work_id":"13d1579e-096c-46e4-a4fd-f45ad09f3101","ref_index":5,"cited_arxiv_id":"1509.05009","is_internal_anchor":true}],"resolved_work":30,"snapshot_sha256":"805623669fa5bd10d7e6be64a87d80d826136b18482a2628b7157f453d4181cb","internal_anchors":11},"formal_canon":{"evidence_count":2,"snapshot_sha256":"9e63ee3cdb2c5dc9ee44e5923f79d58e73d5d776f03c9605e767cf34da3702c1"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"77df47ca-3fa3-4e3d-acf6-45a3b304390e"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T21:57:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XDMK+4Qn4SWvza9T3U083QcWoqK2Mz33chYkAr2alrPIoCmekadOXnyyZTewE4VoMdze+x+Gduo4fqdyb4UdDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-19T00:00:46.461800Z"},"content_sha256":"903262bcd57b56da449f8ff8794d010fbe8ede7c7f11f101985c1979f4657b8d","schema_version":"1.0","event_id":"sha256:903262bcd57b56da449f8ff8794d010fbe8ede7c7f11f101985c1979f4657b8d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6ONAGNNXV7PBJCVS34VTAMMXYU/bundle.json","state_url":"https://pith.science/pith/6ONAGNNXV7PBJCVS34VTAMMXYU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6ONAGNNXV7PBJCVS34VTAMMXYU/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":"2026-05-19T00:00:46Z","links":{"resolver":"https://pith.science/pith/6ONAGNNXV7PBJCVS34VTAMMXYU","bundle":"https://pith.science/pith/6ONAGNNXV7PBJCVS34VTAMMXYU/bundle.json","state":"https://pith.science/pith/6ONAGNNXV7PBJCVS34VTAMMXYU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6ONAGNNXV7PBJCVS34VTAMMXYU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:6ONAGNNXV7PBJCVS34VTAMMXYU","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":"dcce65d5f71ac5fafe019f00c52eff421f7b4aecff078765275eb1bedcc95775","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T17:58:27Z","title_canon_sha256":"7bde740558cb6b4bd2e9b8c8ef41968c551bd3f797b08f522c06d4d8b1fe8847"},"schema_version":"1.0","source":{"id":"2605.15183","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.15183","created_at":"2026-05-17T21:18:32Z"},{"alias_kind":"arxiv_version","alias_value":"2605.15183v1","created_at":"2026-05-17T21:18:32Z"},{"alias_kind":"pith_short_12","alias_value":"6ONAGNNXV7PB","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"6ONAGNNXV7PBJCVS","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"6ONAGNNX","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:903262bcd57b56da449f8ff8794d010fbe8ede7c7f11f101985c1979f4657b8d","target":"graph","created_at":"2026-05-17T21:57:18Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Tensor similarity captures global functional equivalence and accounts for cross-layer mechanisms using an efficient recursive algorithm. This reduces measuring similarity and verifying faithfulness into a solved algebraic problem rather than one of empirical approximation."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the recursive algorithm correctly identifies functional equivalence for all tensor-based models without missing non-linear interactions or symmetries outside weight-space basis changes."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Tensor similarity is a symmetry-invariant metric that measures functional equivalence between tensor-based networks using a recursive algorithm for cross-layer mechanisms."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Tensor similarity is a weight-based metric that algebraically determines when two neural networks implement the same computation by ignoring irrelevant symmetries."}],"snapshot_sha256":"4890af74f90e38a0855cccf7b6724135c0954aa81b1e826dd7cd32d69f70e799"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"9e63ee3cdb2c5dc9ee44e5923f79d58e73d5d776f03c9605e767cf34da3702c1"},"paper":{"abstract_excerpt":"Mechanistic interpretability aims to break models into meaningful parts; verifying that two such parts implement the same computation is a prerequisite. Existing similarity measures evaluate either empirical behaviour, leaving them blind to out-of-distribution mechanisms, or basis-dependent parameters, meaning they disregard weight-space symmetries. To address these issues for the class of tensor-based models, we introduce a weight-based metric, tensor similarity, that is invariant to such symmetries. This metric captures global functional equivalence and accounts for cross-layer mechanisms us","authors_text":"Jacob Meyer Cohen, Laurence Wroe, Logan Riggs Smith, Melwina Albuquerque, ML Nissen Gonzalez, Thomas Dooms","cross_cats":[],"headline":"Tensor similarity is a weight-based metric that algebraically determines when two neural networks implement the same computation by ignoring irrelevant symmetries.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T17:58:27Z","title":"When Are Two Networks the Same? Tensor Similarity for Mechanistic Interpretability"},"references":{"count":30,"internal_anchors":11,"resolved_work":30,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"David Bau, Bolei Zhou, Aditya Khosla, Aude Oliva, and Antonio Torralba","work_id":"823a63dc-e714-4553-a714-37fe028b7fce","year":2025},{"cited_arxiv_id":"1704.05796","doi":"","is_internal_anchor":true,"ref_index":2,"title":"Available: https://arxiv.org/abs/1704.05796","work_id":"4ecff60a-03d5-476d-b6b9-cba59c37a41c","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Neural networks learn statistics of increasing complexity","work_id":"3803730c-4a5c-4f3e-aafc-3ff58e1860a6","year":null},{"cited_arxiv_id":"1603.00162","doi":"","is_internal_anchor":true,"ref_index":4,"title":"Convolutional Rectifier Networks as Generalized Tensor Decompositions","work_id":"bc48707b-0c47-4ea8-91b0-97b76c7d0dbe","year":null},{"cited_arxiv_id":"1509.05009","doi":"","is_internal_anchor":true,"ref_index":5,"title":"On the Expressive Power of Deep Learning: A Tensor Analysis","work_id":"13d1579e-096c-46e4-a4fd-f45ad09f3101","year":null}],"snapshot_sha256":"805623669fa5bd10d7e6be64a87d80d826136b18482a2628b7157f453d4181cb"},"source":{"id":"2605.15183","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-15T03:13:32.984659Z","id":"77df47ca-3fa3-4e3d-acf6-45a3b304390e","model_set":{"reader":"grok-4.3"},"one_line_summary":"Tensor similarity is a symmetry-invariant metric that measures functional equivalence between tensor-based networks using a recursive algorithm for cross-layer mechanisms.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Tensor similarity is a weight-based metric that algebraically determines when two neural networks implement the same computation by ignoring irrelevant symmetries.","strongest_claim":"Tensor similarity captures global functional equivalence and accounts for cross-layer mechanisms using an efficient recursive algorithm. This reduces measuring similarity and verifying faithfulness into a solved algebraic problem rather than one of empirical approximation.","weakest_assumption":"That the recursive algorithm correctly identifies functional equivalence for all tensor-based models without missing non-linear interactions or symmetries outside weight-space basis changes."}},"verdict_id":"77df47ca-3fa3-4e3d-acf6-45a3b304390e"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:4a11b3d8d76acddbc7de8ee2451cce4f2120bb639cbe39da40caefdc4ba78b18","target":"record","created_at":"2026-05-17T21:18:32Z","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":"dcce65d5f71ac5fafe019f00c52eff421f7b4aecff078765275eb1bedcc95775","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T17:58:27Z","title_canon_sha256":"7bde740558cb6b4bd2e9b8c8ef41968c551bd3f797b08f522c06d4d8b1fe8847"},"schema_version":"1.0","source":{"id":"2605.15183","kind":"arxiv","version":1}},"canonical_sha256":"f39a0335b7afde148ab2df2b303197c5020b979c1c18e65cd4daded5eddefac7","receipt":{"builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f39a0335b7afde148ab2df2b303197c5020b979c1c18e65cd4daded5eddefac7","first_computed_at":"2026-05-17T21:40:25.131488Z","kind":"pith_receipt","last_reissued_at":"2026-05-17T21:57:18.510566Z","receipt_version":"0.2","signature_status":"unsigned_v0"},"source_id":"2605.15183","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4a11b3d8d76acddbc7de8ee2451cce4f2120bb639cbe39da40caefdc4ba78b18","sha256:903262bcd57b56da449f8ff8794d010fbe8ede7c7f11f101985c1979f4657b8d"],"state_sha256":"e3ce5819ce206c94ec6b37bc1a40141cc909527bc64992498371435ba9ba847c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PaIa4Rpjq11MXyp0DCpCynrXgbATUTD9dtW4GZ7c1PA4SIm17Bn7GLduKSG+OCn11FSvbqjKNptx3BbGoc6kCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-19T00:00:46.464869Z","bundle_sha256":"60f812f12ba29649fde5d61ff395c4e470193cfb62da078664eb026cc20ee47a"}}