3
[2026-09-23] · src LHCb Collaboration; arXiv:2609.25178 (2026-09-21)
Reconstructing Bs0 mesons in three decay modes (Bs0->Ds-pi+, Bs0->J/psi phi, Bs0->Ds-pi+pi+pi-) and combining each with a pi0, LHCb finds a peaking structure in the Bs0pi0 invariant-mass spectrum at 5698.9 +/- 1.5 (stat) +/- 0.6 (syst) +/- 0.1 MeV/c2, with global significance exceeding 7 sigma and a natural width below 9.8 MeV at 90% CL (11.8 MeV at 95% CL). Both the mass - well under conventional quark-model predictions for the lightest P-wave Bs0 excitation - and the narrowness are what the paper calls unexpected: the state is consistent either with the long-sought chiral partner of the ground-state Bs0 (the "chiral doubling" pattern already seen in the D and Ds systems) or with the first exotic multiquark hadron found in the open single-beauty system, and the paper does not adjudicate between the two.
Chiral partners of heavy-light mesons have been predicted for two decades and confirmed only in the charm sector; a beauty-sector partner this light would extend that pattern to a regime where the light-quark dynamics should, in principle, decouple further from the heavy quark, which is itself informative about how universal the mechanism is. The alternative reading - a genuine exotic state - would be the first of its kind outside the charm and doubly-charm tetraquark/pentaquark family that LHCb has spent a decade cataloguing, opening a beauty-sector analogue nobody has confirmed before. Which of the two is true is not yet decided by this paper, and that ambiguity is itself the finding: a single narrow, light state that current theory cannot cleanly assign to either category.
[2026-09-23] · src Super-Kamiokande Collaboration; arXiv:2609.25681 (2026-09-22)
Searching 0.401 megaton-years of Super-Kamiokande exposure across its five run periods (SK-I through SK-V, 1996-2020) for the |Delta B|=3 decay O-16(ppp) -> C-13 pi+ pi+ e+ - three bound protons converting simultaneously - the collaboration finds no signal and sets a lower limit on the partial lifetime of 4.2x1032 years, an improvement of six orders of magnitude over the previous experimental constraint. The gain comes chiefly from a convolutional neural network with an attention mechanism classifying Cherenkov-ring event topology, the first time a CNN has been applied to a nucleon-decay search at Super-K.
Most nucleon-decay searches target |Delta B|1 modes (p -> e+ pi0 and its relatives), which probe GUT-scale gauge-boson exchange directly. |Delta B|3 operators are a different, higher-dimension window: they matter for baryogenesis scenarios built to survive electroweak sphaleron washout, which erases some lower-Delta-B asymmetries but not all of them, so a bound here constrains model space the standard proton-decay channels do not touch. Six orders of magnitude in one step is a large jump for a search that had gone essentially untouched by modern analysis techniques; it is as much a demonstration that decades-old Super-K data still has headroom for a machine-learning re-analysis as it is a new number for any specific GUT model.
[2026-09-23] · src Rea, et al.; arXiv:2609.25990 (2026-09-22); companion IXPE polarimetry: Taverna, et al.; arXiv:2609.25958
The Einstein Probe Wide-field X-ray Telescope caught this magnetar early in an outburst on 2026-06-28, and rapid multi-instrument follow-up (XMM-Newton, NuSTAR, SVOM, IXPE) tracked its luminosity falling by roughly half within the first month while confirming the identification via short X-ray bursts. The measured spin period (6 s), dipolar field (2.6x1014 G) and characteristic age (~34 kyr) are unremarkable for a magnetar, but its position - Galactic latitude b=-4.56 deg, about 0.26 kpc below the midplane - is not: reconciling that height with an age this young requires an extreme proper-motion velocity under a simple midplane-birth-and-kick picture. No infrared or radio counterpart was found despite sensitive searches.
This is a formation-site puzzle rather than a new physical mechanism: either the characteristic age (a spin-down estimate, not a direct clock) is misleading here, the natal kick was unusually large, or the birth site itself was already off the midplane - each of which has different implications for how magnetars end up distributed through the Galaxy. On its own this is one object, not a population result, but it is also a demonstration of what Einstein Probe's wide-field monitoring is for: catching quiescent magnetars at outburst onset, before selection effects from older, colder surveys shape the sample. More examples like this one, not this one alone, are what would turn the puzzle into an answer.
2
[2026-09-23] · src twelve-keys/src/exercises/blues_rootless/exercise.py
Piano gets the hour today: three days since its last session on the 20th is the longest real gap of the three instruments right now, alto went yesterday and trumpet the day before, continuing the alto/piano/trumpet cycle this rotation has held since mid-September. It opens lap three's blues slot with the Evans A-form rootless voicing — a four-note 3-13-b7-9 grip with no root, the same rootless-shape vocabulary 09-14's shells session began, now applied to an actual twelve-bar form for the first time rather than drilled as a shape in isolation. Key is C, the last of the twelve still untouched by this panel, and the one that strips away every black-key landmark the hand has been leaning on through three sessions moving off the flat side.
5 minutes: left hand alone, the C7 grip (E-A-Bb-D) held as a whole note, then the F7 grip (A-D-Eb-G), each named aloud bottom-up — "three, thirteen, flat seven, nine" — before playing it. The point is locating each note by name and feel before any tempo is asked of the hand.
Hands together, quarter = 80, six consecutive full twelve-bar choruses with no stop between them: LH holds each bar's voicing as a whole note following the printed form (I IV I I / IV IV I I / V IV I V), RH adds the guide tones written above it. Success criterion: on every I-to-IV move — bar 1 into bar 2, and again bar 4 into bar 5 — the two fingers on A and D stay physically still, and only the other two move, E sliding down to Eb and Bb dropping to G. A pass where those two fingers lift and reset doesn't count toward the six.
One chorus, shaped as a single arc rather than twelve identical bars: the RH guide-tone line sits on E-Bb for most of the form and only lifts to B-F twice, bar 9 and the closing turnaround at bar 12. Lean into both lifts — a little more volume, a little more space before the next attack — so the one harmonic event the form actually has reads as an event rather than passing unmarked.
Stop the moment the left hand starts hunting for a grip instead of finding it by feel — a rootless voicing that needs a glance down isn't internalized yet, and reps played while searching train the search, not the shape.
[2026-09-23] · src —
hearCharlie Parker's Reboppers — Billie's Bounce · Tidal · Spotify
Today's exercise names its own application tunes, and this is the closest of the three: an F blues built on the same quick-change form printed above the C exercise, I moving to IV already in the second bar rather than staying put for four. Cut 26 November 1945 for Savoy under Parker's own name, with Miles Davis on trumpet, Curley Russell on bass and Max Roach on drums. Listen past the horns to what the harmony is doing under them on that first change — the same two-chord pivot today's technical target is drilling, at the tempo and in the setting the shape was actually built for.