Metraipolis › archive › dc_3758aab4a5fe

Causal Structure is Inducible but Functionally Decoupled: The Routing/Readout Boundary of a Typed Mechanism Library — arXiv:2608.11767

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.11767
License (as archived)
CC0-1.0
Archived
2026-10-19T05:27:35.891Z
SHA-256 of the archived text
b7150da1f26440335156d1b0269d14bbd59051fe70a1a9fb1d2d3818796a0b8b
Stable URL of this page
https://metraipolis.com/documento/dc_3758aab4a5fe

How to cite

Causal Structure is Inducible but Functionally Decoupled: The Routing/Readout Boundary of a Typed Mechanism Library — arXiv:2608.11767 — arXiv (en), https://arxiv.org/abs/2608.11767 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3758aab4a5fe (Metraipolis archive dc_3758aab4a5fe, sha256 b7150da1f26440335156d1b0269d14bbd59051fe70a1a9fb1d2d3818796a0b8b)
BibTeX
@misc{dc3758aab4a5fe,
  title = {Causal Structure is Inducible but Functionally Decoupled: The Routing/Readout Boundary of a Typed Mechanism Library — arXiv:2608.11767},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.11767},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3758aab4a5fe; sha256 b7150da1f26440335156d1b0269d14bbd59051fe70a1a9fb1d2d3818796a0b8b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.11767. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

When a language model answers an interventional question, the computation it must perform depends on the type of evidence the query requires. We report a decoupling in how a transformer organizes causal knowledge: slot-by-type structure induced by type-level supervision organizes routing, yet remains functionally decoupled from answer readout. We establish this with a typed mechanism library -- discrete mechanism slots partitioned by evidence type, auditable at the state level -- on a causal-world benchmark with exact interventional ground truth, under a frozen protocol, at two scales (22.6M and 125M). Four preregistered findings. (i) Origin. Slot-by-type organization is induced by type-level supervision: absent in architecturally identical unsupervised controls, not buyable by content-free gating labels, and statistically attributable to the supervision signal, replicating at 125M under a powered preregistered protocol (all nine cells passed). (ii) Boundary. The induced structure is a typed routing index with a sharp routing/readout boundary: slot codes scaffold routing but do not drive answer readout ($|\Delta\hat{y}| \le 3.4\times10^{-6}$, zero collateral, three seeds, stable across a 5.6x scale window) -- we therefore make no behavioral-editability claim. (iii) Cost. The structure is free: LM quality matches a parameter-matched monolith within 0.0082 nats. (iv) Trust. The library state is exactly local under edit and bit-exactly revertible -- 250 single-edit and 1,000 stacked reverts per seed, zero failures. We further find that the unsupervised null itself moves with scale, so comparisons reusing a null calibrated at one scale may be confounded at another. Every claim is tied to a preregistered, machine-checkable criterion archived before the data it governs; the full audit trail, including one criterion we failed and how the frozen protocol handled it, is released as an appendix.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-13 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.11767 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T05:15:39.212Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-13?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).