Explanation statuses — Whether the record believes the ACCOUNT, which is not what the same word means on a practice. An explanation can be Rejected while the thing it explained goes on working exactly as well as it ever did (ADR-031).
Disbelieved — tested and failed, or contradicted by later work — and the thing it explained may still work perfectly well.
5 of 1352 THEORY documents. Back to the full index.
| # | Title | Summary | Status |
|---|---|---|---|
| THEORY-024 v2 | Classical Boolean logic is restored when a presheaf of contextual data acquires a global section or is sheafified | Ghose (2025), LIT-276 — asserts that the Heyting algebra of truth values “becomes Boolean” when the presheaf data collapse to a global section, and that sheaves encode Boolean reasoning. Both claims are asserted, not proved. The record’s readings refute them. The internal logic of a topos is fixed by the base category and its topology, not by whether one presheaf has a section (Isham–Butterfield, LIT-325, §6). Sheaf toposes such as Sh(ℝ) are not Boolean (NOTE-249). The contextuality criterion itself is untouched. | Rejected |
| THEORY-035 v5 | Deep networks generalise because the diffusion phase of SGD compresses each layer’s information about the input | Shwartz-Ziv & Tishby (2017), LIT-324, report a fitting phase followed by a long compression phase in the information plane, and read compression by gradient diffusion as the mechanism of generalisation. Their §3.4 reports that with 5% of the data compression lowers I(T;Y). Saxe et al. (2018), LIT-372, rerun their code and find the compression phase only in double-saturating networks measured through an imposed binning, find it under full-batch GD as under SGD, and find compression and generalisation in all four combinations. The account is rejected. The IB functional of LIT-338, the tanh two-phase observation and the gradient-SNR transition are not. | Rejected — Saxe et al. show the compression phase appears only for double-saturating units under an imposed binning or noise model, occurs under full-batch GD as under SGD, and dissociates from generalisation in all four combinations; the paper’s own small-sample run already showed compression losing label information |
| THEORY-765 v2 | Brushing a seen rubber hand in synchrony with the hidden real hand makes touch be felt at the rubber hand, makes it be reported as one’s own, and shifts the felt position of the real hand toward it; a small asynchrony removes these effects | Botvinick & Cohen (1998), LIT-1277, read in full in NOTE-1006. Ten subjects affirmed referred touch and ownership after ten minutes of synchronous brushing; after thirty minutes their blind reaches drifted toward the rubber hand in proportion to how long the illusion was reported; asynchronous brushing cut both. It does not show that synchrony alone, apart from expectation, produces the report, and it concerns a hand, not the whole body or the self. | Rejected — Rejected on 2026-10-20 by its own refutation condition: drift toward the rubber hand appeared as strongly under asynchronous brushing interrupted by pauses, and with the hand only seen (Rohde et al. 2011, LIT-1281). Its suggestion control was also too weak to separate perception from expectation (Lush & Seth 2022, LIT-1267). The report half is restated, narrower, in THEORY-766. |
| THEORY-787 v2 | A system’s causal structure can be most effective at a macro scale, so that causation emerges at a coarse-grained level and the macro description is the one at which the system’s causes are found, as Hoel, Albantakis and Tononi propose from effective information | Hoel, Albantakis and Tononi (2013), LIT-1038, argue that macro can beat micro: the scale at which effective information peaks is where a system’s causal structure is, so causation can emerge at a macro level. The record holds as shown, in THEORY-222, only the arithmetic: under uniform interventions imposed separately at each level, a macro description can have more effective information. This is the causal thesis, filed under ADR-047 v4 when the narrowing dropped it. | Rejected — Rejected on 2026-10-30, at the owner’s direction, by its own refutation condition: the macro advantage disappears whenever the same intervention distribution over micro states is used at both levels. THEORY-1167 proves this for a coarse-graining of one transition, by data processing, and Comolatti and Hoel concede it (LIT-1843). The independent interventionist test the promote_when asks for, causal feature learning (LIT-1833), can confirm that a macro is proportionate but never that it beats the micro (THEORY-1169). What the record holds instead is CLAIM-497, by ARG-041. Hoel’s temporal and spatio-temporal macros, which the bound does not reach, are QUESTION-335. The arithmetic it rested on, THEORY-222, stands. |
| THEORY-1011 v3 | In autoregressive transformer language models, factual associations are stored in middle-layer MLP modules as linear key–value memories: the MLP takes a key encoding the subject at its last token and returns a value encoding its attributes, which late attention moves to the prediction, and any middle layer could store a given fact equally, as Meng, Bau, Andonian and Belinkov propose | Meng et al. (2022), LIT-1536, read in NOTE-1166, §2.3 and §3: “the localized factual association hypothesis”. It localizes recall along three dimensions, “(i) in the MLP modules (ii) at specific middle layers (iii) and specifically at the processing of the subject’s last token”, and conjectures “that any fact could be equivalently stored in any one of the middle MLP layers”. The support is causal tracing and the success of rank-one edits at the traced site. | Rejected — Hase et al. (LIT-1578, THEORY-1024) meet this entry’s own refutation clause for its localization part, in GPT-J on 652 CounterFact facts. ROME rewrites a fact at over 96% at every layer but the last, and the per-fact tracing effect adds at most 3.2 points of R² to edit success. The conjecture that any middle layer could store a fact equally was split out to THEORY-1039 and stays Proposed |