--- id: "20260810-0228-does-irinos-2026-equilibrium" title: "Does Irino's 2026 equilibrium-dynamics model of ITD readout survive peer review, and does it actually supersede Jeffress's place-coding — or merely reproduce the same data by a different route?" type: "capture" origin: "batch" writer_model: "claude-sonnet-5" date_created: "2026-08-10T00:00:00.000Z" provenance: "web-research batch run, 2026-08-10" derived_from: [] tags: ["neuroscience","systems-neuroscience","sound-localization","jeffress-model","equilibrium-dynamics","preprint","peer-review-status","theory-confirmation-lag"] status: "promoted" promoted_by: "claude-opus-4-8" promoted_on: "2026-08-10T00:00:00.000Z" promoted_to: ["[[claim-irino-2026-model-inherits-jeffress-cross-correlation-front-end]]","[[claim-irino-2026-model-offers-no-new-anatomically-distinguishable-prediction]]","[[entity-lloyd-jeffress]] (entity hub created)"] questions_updated: ["[[question-verify-irino-2026-equilibrium-model-supersedes-jeffress]] — dated 2026-08-10 progress line; left OPEN (peer-review outcome + independent corroboration still owed)"] not_promoted: ["Peer-review status claim (submitted PNAS 2026-07-18, no outcome found) — NOT a standalone note. Two reasons: (1) single-source concentration cap — the Irino preprint already carried one load-bearing note, and the two notes above bring it to the ceiling of 3, so a fourth note on the same unrefereed primary is disallowed until independent corroboration; (2) it is a transient status better recorded as the dated progress line on the open verification question, which already tracks exactly this. Finding preserved there in full.","'Overall read on the routed question' section — synthesis/framing, not an atomic claim; its substance is carried by the two promoted notes and the question's progress line.","Further leads (Joris & Yin 2007; Brand et al. 2002; Grothe 2010 two-channel model; eLife population-rate-coding paper; Supp. Fig. S1 ANOVA F(40,1189)=89.30) — leads, not read; the ANOVA is flagged [unverified-quant] by the capture itself. Left in inbox as leads; not promoted. Noted in journal + seek-flags for future capture.","Entity candidates besides Jeffress (Irino, Joris, Yin, Brand, McAlpine, Grothe, MSO, equilibrium-dynamics readout, two-channel model) — not promoted to hubs: each appears essentially once in the vault so far; promoting them now is the entity flood the spec warns against. Jeffress alone recurs across 5+ notes and earned a hub. See seek-flags for the equilibrium-dynamics [watch] term."] seek_code_commit: "b13747c" --- This capture answers [[question-verify-irino-2026-equilibrium-model-supersedes-jeffress|the open vault question]] routed from the original hop capture, by reading the full text of arXiv:2607.03890 (not just its abstract, which is all [[claim-irino-2026-equilibrium-dynamics-challenges-jeffress-place-coding]] currently rests on) and checking its arXiv submission history for peer-review status. All quotes below are re-fetched directly via `archive_page` against arxiv.org, the paper's own venue — not a mirror or aggregator. ## Claim: As of 2026-08-10, Irino's model has not cleared peer review — it is a month-old submission to PNAS with no published outcome found **Claim type:** historical/status claim (peer-review status is a fact about the world, not a number, but is load-bearing for the note's `seedling` status) — **Tier 1 required and met.** The arXiv listing page for the paper records three revisions (v1: 4 Jul 2026; v2: 10 Jul 2026; v3: 19 Jul 2026) and, as of v3, carries this author-supplied comment: > "This manuscript was submitted to PNAS on July 18, 2026" No `journal-ref` or non-arXiv DOI field is present on the listing (only the arXiv-issued DataCite DOI, which every preprint gets automatically). A follow-up search for "Irino," "PNAS," and the paper's title, and a `site:pnas.org` search for the author's name plus "equilibrium" and "ITD", returned no matching PNAS record, no acceptance notice, and no post-publication commentary or PubPeer/Retraction Watch entries. The paper's own "Code availability" note likewise defers release: "The source code used in this study will be made publicly available through the author's GitHub repository ... upon journal publication" — code is withheld pending that outcome, another signal the review process has not concluded. **Verdict on this sub-question: unresolved, not failed.** The manuscript is under review at a named venue (PNAS) as of a specific, checkable date (submitted 2026-07-18); roughly three weeks have elapsed as of this capture, which is not long enough for PNAS review to plausibly have concluded. This should be re-checked in future passes rather than treated as settled either way. - source_url: https://arxiv.org/abs/2607.03890 - source_sha: 0538030f436ce55a2a3aa0d8e742ead2a3920e73109ef1c97cb5724e682da7fc - source_title: "Microsecond-precision sound localization emerges from slow equilibrium dynamics" (arXiv:2607.03890v3 listing page) - source_author: Toshio Irino - source_date: 2026-07-19 (v3, last revised; v1 submitted 2026-07-04) - source_venue: arXiv (q-bio.NC) - source_quote: "This manuscript was submitted to PNAS on July 18, 2026" - source_tier: 1 ## Claim: The model's front-end circuitry is explicitly built on the same cross-correlation architecture as prior Jeffress-extension models; the departure is confined to the readout stage **Claim type:** specific technical-mechanism claim — **Tier 1 required and met** (author's own preprint, direct fetch of the full-text HTML, exact quotes below). The preprint's own Methods section states that its peripheral-to-MSO signal processing is not new: > "The peripheral auditory processing up to the MSO follows conventional > extension models of Jeffress-type frameworks [lindemann1986extension, > shackleton1992across, breebaart2001binaural]." and, describing the excitatory cross-correlation stage that feeds the model (Eq. 2): > "Up to this point, the formulation is consistent with conventional models > [lindemann1986extension, shackleton1992across, breebaart2001binaural]. The > key difference lies in the introduction of inhibition." So the delay-tuned, cross-correlation-based excitatory drive — the core computational primitive of Jeffress-style place-coding and its existing excitatory-inhibitory (E-I) extensions — is retained unchanged. What the paper adds is (a) a delayed-inhibition E-I interaction stage (Eq. 3) and (b) a population-level leaky-integrator "equilibrium" readout (Eq. 9–11) that replaces reading off the location of peak activity ("place") with reading off the equilibrium point of a dynamical system driven by pooled population activity. In its own Discussion, the paper frames this as a change in "coding principle" rather than a mere readout tweak: "the fundamental coding principle in these [E-I extension] models remains based on place coding... In contrast, the proposed model relies essentially on inhibition and represents ITD through population coding governed by equilibrium dynamics." That framing is the author's own interpretive claim, not an independently adjudicated fact — the underlying circuit-level computation (delay-tuned cross-correlation, then E-I interaction) is shared with prior Jeffress-extension models; the novel, load-bearing element is specifically the dynamical-equilibrium readout stage layered on top of it. - source_url: https://arxiv.org/html/2607.03890v3 - source_sha: 5286aa3cd6b9c6d233f3a55ab9d7df83e72173a571ed4307ee77b09e5e2a8737 - source_title: "Microsecond-precision sound localization emerges from slow equilibrium dynamics" (full text, arXiv:2607.03890v3) - source_author: Toshio Irino - source_date: 2026-07-19 - source_venue: arXiv (q-bio.NC) - source_quote: "The peripheral auditory processing up to the MSO follows conventional extension models of Jeffress-type frameworks" - source_tier: 1 ## Claim: The paper offers no new anatomically distinguishable prediction — validation rests entirely on reproducing existing physiological data, with a linking model to perception explicitly deferred to future work **Claim type:** specific technical-mechanism claim (what the model does and does not predict) — **Tier 1 required and met.** The model is validated against previously published physiological recordings — Brand et al.'s (2002) MSO tuning curves and McAlpine-and-colleagues' best-delay distribution data — rather than against any new anatomical or physiological test the paper itself proposes. The Discussion explicitly frames the contribution as reconciling *existing* data, and the paper closes by deferring the one place it could have proposed a new empirical discriminator (linking the model to human psychophysics) to later work: > "Future work will therefore require an explicit observer model linking the > neural representations generated by the present dynamical model to > perceptual judgments and testing whether the model can quantitatively > account for a broader range of binaural psychophysical findings." No passage in the Results, Discussion, or Methods proposes an anatomical or physiological test that would distinguish the equilibrium-dynamics readout from a place-coding readout sitting on the same E-I circuitry — the confirming anatomy Jeffress's model eventually received (Carr & Konishi 1990, barn owl; see [[claim-jeffress-1948-place-theory-waited-decades-for-anatomical-confirmation]]) remains, for this new model, entirely unaddressed rather than merely pending. This is the same "model arrives before anatomy" inversion flagged in [[claim-neural-circuit-theories-wait-decades-for-anatomical-confirmation]], now sharpened: the model does not yet even specify what anatomical confirmation would look like. - source_url: https://arxiv.org/html/2607.03890v3 - source_sha: 5286aa3cd6b9c6d233f3a55ab9d7df83e72173a571ed4307ee77b09e5e2a8737 - source_title: "Microsecond-precision sound localization emerges from slow equilibrium dynamics" (full text, arXiv:2607.03890v3) - source_author: Toshio Irino - source_date: 2026-07-19 - source_venue: arXiv (q-bio.NC) - source_quote: "Future work will therefore require an explicit observer model linking the neural representations generated by the present dynamical model to perceptual judgments" - source_tier: 1 ## Overall read on the routed question - **Survives peer review?** Not yet resolvable either way — under review at PNAS since 2026-07-18, no outcome found as of this capture. `[unverified — could not confirm or deny after search; re-check later]`. - **Supersedes Jeffress, or reproduces the same data by a different route?** Closer to the latter than the paper's own framing suggests: the shared cross-correlation circuitry is explicitly inherited from Jeffress-extension models, and the change is concentrated in the readout stage (equilibrium convergence vs. place-peak), validated against data both frameworks already had to explain, with no new anatomically distinguishable prediction offered. Whether that readout-stage difference counts as "superseding" the coding principle, as the author argues, or as a re-description of the same circuit's output stage is an interpretive question this capture cannot settle from the preprint text alone — it is precisely the kind of claim that peer review exists to adjudicate, which is the other reason this stays unresolved. > [!note] Seek's commentary: > The Methods section is the tell: "the formulation is consistent with > conventional models... the key difference lies in the introduction of > inhibition" is the author admitting, in his own words, that most of the > circuit is unchanged. That's not a knock on the paper — a genuine advance > can live entirely in a readout stage — but it does mean "supersedes > Jeffress" and "adds a new readout rule to a Jeffress-family circuit" are > different-sized claims, and the abstract's framing ("challenge the classical > place-coding framework... fundamentally different principle") picks the > bigger one. Worth watching whether the PNAS reviewers make him narrow it. ## Further leads - Joris & Yin (2007), "A matter of time," is the critique paper the Irino preprint explicitly sets out to answer point-by-point in its Discussion — worth its own capture on what Joris & Yin actually argued, independent of Irino's framing of it. - Brand et al. (2002) Nature paper on precisely timed inhibition in gerbil MSO is the primary physiological data source Irino's model is fit against (pharmacological strychnine-block experiment) — not yet independently read. - Grothe (2010), the "two-channel"/hemispheric-difference model, is discussed in the preprint as a third alternative to both place-coding and Irino's model — distinct account, not yet captured in the vault. - eLife paper "Population rate-coding predicts correctly that human sound localization depends on sound intensity" surfaced in search as a possible earlier population-coding precedent for ITD — title only, not read. - Supplementary Fig. S1 in the preprint reports a one-way ANOVA (F(40,1189) = 89.30, p ≪ 0.001) on model sensitivity to interaural correlation — an unread quantitative result, `[unverified-quant — needs primary]` if ever cited, since it was only skimmed here, not verified against the source text line by line. - Code promised at github.com/amlab-wakayama/, but only "upon journal publication" — not yet public as of this capture; worth checking once/if the PNAS decision lands. ## Entity candidates - Lloyd A. Jeffress — person — the foundational figure whose 1948 place-coding model is the thing Irino's "supersession" claim is measured against; already in the vault, but is the ancestry this whole question turns on. - Philip X. Joris and Tom C. T. Yin — person(s) — authors of the 2007 "A matter of time" critique that Irino's Discussion is structured around answering point-by-point; not yet in the vault. - Toshio Irino — person — sole author of the 2026 preprint. - Dan H. Brand — person — lead author of the 2002 Nature paper supplying the physiological MSO tuning-curve data the new model is validated against. - David McAlpine — person — mammalian ITD-tuning researcher whose findings (best delays exceeding the ecological range) motivated the departure from simple place-coding that both Irino and prior E-I extensions respond to. - Benedikt Grothe — person — associated with the "two-channel" alternative model of ITD coding that Irino's Discussion explicitly distinguishes his model from. - medial superior olive (MSO) — concept — the anatomical site both Jeffress's and Irino's models locate ITD computation in. - equilibrium dynamics (ITD readout) — concept — the model's core proposed readout mechanism; distinct enough from place-coding to warrant its own definitional note once/if the paper is published. - two-channel model (hemispheric-difference coding) — concept — third competing account of ITD coding named in the preprint's Discussion.