{"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-2601-03424","title":"Spectral Archaeology: The Causal Topology of Model Evolution","arxiv_id":"2601.03424","date":"2026-01-06","proceeding":null,"authors":["Valentin Noël"],"abstract":"Behavioral benchmarks tell us \\textit{what} a model does, but not \\textit{how}. We introduce a training-free mechanistic probe using attention-graph spectra. Treating each layer as a token graph, we compute algebraic connectivity ($λ_2$), smoothness, and spectral entropy. Across 12 models and 10 languages, these measures yield stable ``spectral fingerprints'' that expose discontinuities missed by standard evaluation. We report four results. (1) Models undergoing specific curriculum transitions (e.g., code-to-chat) show an English-only, syntax-triggered connectivity failure on non-canonical constructions, reaching $Δλ_2 \\approx -0.76$. We term this scar \\textit{Passive-Triggered Connectivity Collapse} (PTCC). Analysis of the Phi lineage reveals that PTCC appears and resolves across developmental stages, implicating brittle curriculum shifts rather than synthetic data per se. (2) PTCC reflects a specialization trade-off: strengthened formal routing at the expense of stylistic flexibility. (3) We identify four recurrent processing strategies; simple frozen-threshold rules enable perfect forensic identification across lineages. (4) Mechanistically, PTCC localizes to a sparse Layer 2 ``compensatory patch'' of heads that fails under syntactic stress; activation steering can partially restore connectivity, recovering $\\approx 38\\%$ of lost information flow. Finally, dominant topological regimes track tokenization density more than language identity, suggesting ``healthy'' geometry varies systematically across scripts. Overall, attention-graph spectra provide a practical tool for auditing and training-regime verification.","url_abs":"https://arxiv.org/abs/2601.03424","url_pdf":"https://arxiv.org/pdf/2601.03424","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":{"atlas_url":"https://app.syntology.ai/?focus=2601.03424","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2601.03424"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/vcnoel/spectral-archeology","reach":{"status":"ok","spdx":"AGPL-3.0"}}],"summary":{"unverified":6},"by_repo_kind":{"found_in_text":{"samples":6,"ran":0,"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":6,"samples":[{"code_sha256_prefix":"3431b9778164a254","entry":"extract_metric","repo":"vcnoel/spectral-archeology","repo_kind":"found_in_text","path":"src/plotting.py","file_url":"https://github.com/vcnoel/spectral-archeology/blob/HEAD/src/plotting.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"AGPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"3431b9778164a254"}},{"code_sha256_prefix":"b0e62e41d171a975","entry":"generate_pairs","repo":"vcnoel/spectral-archeology","repo_kind":"found_in_text","path":"temp_view.py","file_url":"https://github.com/vcnoel/spectral-archeology/blob/HEAD/temp_view.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"AGPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"b0e62e41d171a975"}},{"code_sha256_prefix":"efa5c41a0e4040ad","entry":"get_fiedler_value","repo":"vcnoel/spectral-archeology","repo_kind":"found_in_text","path":"temp_view.py","file_url":"https://github.com/vcnoel/spectral-archeology/blob/HEAD/temp_view.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"AGPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"efa5c41a0e4040ad"}},{"code_sha256_prefix":"15460bdbf5b5b957","entry":"load_dataset","repo":"vcnoel/spectral-archeology","repo_kind":"found_in_text","path":"src/analysis.py","file_url":"https://github.com/vcnoel/spectral-archeology/blob/HEAD/src/analysis.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"AGPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"15460bdbf5b5b957"}},{"code_sha256_prefix":"66a262d0a1256171","entry":"register_ablation_hooks","repo":"vcnoel/spectral-archeology","repo_kind":"found_in_text","path":"src/analysis.py","file_url":"https://github.com/vcnoel/spectral-archeology/blob/HEAD/src/analysis.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"AGPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"66a262d0a1256171"}},{"code_sha256_prefix":"6cda92e637f8e5bd","entry":"sample_inputs","repo":"vcnoel/spectral-archeology","repo_kind":"found_in_text","path":"src/analysis.py","file_url":"https://github.com/vcnoel/spectral-archeology/blob/HEAD/src/analysis.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"AGPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"6cda92e637f8e5bd"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}