3
[2026-09-17] · src ATLAS Collaboration; arXiv:2603.26463; Phys. Rev. Lett. 137, 111804 (2026-09-11)
Using H -> ZZ* -> l+l-l+l- events from 13 and 13.6 TeV LHC proton-proton collisions, ATLAS measured the angular observables that fix the ZZ spin-density-matrix elements C2,1,2,-1 and C2,2,2,-2, finding -0.71 +/- 0.45 and 0.08 +/- 0.44 respectively - both consistent with the Standard Model. The point of the measurement is not those two numbers on their own but a hypothesis test built from the full angular distribution: it rejects the hypothesis that the two Z bosons' spins are in a separable (unentangled) state at 4.7 standard deviations, against an expected sensitivity of 4.9 sigma if the Standard Model's entangled prediction is correct. This is the first entanglement measurement performed on a pair of massive spin-1 particles - previous LHC entanglement work (top-quark pairs) used spin-1/2 fermions, where the entangled state lives in a two-dimensional qubit space; a Z boson's spin is a three-level qutrit, so the entanglement structure being tested here is genuinely different and richer.
Collider experiments have spent the last few years establishing that LHC data can support real quantum-information analysis - state tomography, Bell-inequality-style tests - rather than just particle counting, but every measurement so far has been on fermion pairs. Massive gauge bosons were an open case: it was not obvious the same angular-observable technique would have the statistical power to see entanglement in a qutrit system extracted from a rare four-lepton final state, and 4.7 sigma against a 4.9 sigma expectation says it does. That opens the same technique to WW and WZ production, where anomalous quartic gauge couplings or other beyond-Standard-Model effects would show up first as a departure from the SM's entanglement prediction rather than as a resonance - a second, largely independent channel for that kind of search on data the LHC is already collecting.
[2026-09-17] · src Balzan, Bera, James, Meyers, Deller, Shannon, Wang; arXiv:2609.18086 (2026-09-16); CRAFT survey (ASKAP)
FRB 20250607A, detected by the CRAFT survey on ASKAP, has two components whose scattering timescale and scintillation bandwidth cannot come from a single thin scattering screen - together they require the dominant screen to sit within roughly 61 kpc of the burst's source, essentially local to it rather than somewhere along the line of sight through the host galaxy or Milky Way. Through that scattering tail, the polarization angle should flatten as the pulse mixes with scattered light; here it does not, a behavior the authors say has not previously been reported in any FRB or pulsar. The two components also show an apparent, visually clear difference in dispersion measure that is nonetheless not statistically significant given current uncertainties - a discrepancy the paper reads as a sign that the uncertainties themselves are too conservative rather than that the difference is illusory. Two further CRAFT-detected bursts, FRB 20190611B and FRB 20210407E, show comparable component-to-component electron-column contrasts, at a level consistent with plasma density variations inside a neutron-star magnetosphere; FRB 20210407E separately shows its own component-dependent polarization-angle structure, and none of the three show generalized Faraday rotation.
Most FRB propagation effects - dispersion, scattering, rotation measure - are usually read as telling you about the matter between the burst and Earth: the host galaxy's interstellar medium, the intergalactic medium, the Milky Way's halo. This paper's case is that for these three bursts, the more consistent reading puts the relevant plasma inside the source's own magnetosphere instead, which is a different regime entirely - dense, ordered, and evolving on the burst's own timescale rather than static along a sightline. If that holds up across more multi-component bursts, it turns FRBs with resolved substructure into a direct probe of the immediate environment of whatever compact object is producing them, rather than only a background light source for studying everything in front of it.
[2026-09-17] · src Chen, Gao, Huang, Tang, Fan; arXiv:2609.17936 (2026-09-15)
Combining GW170817 with an enlarged sample of neutron stars that now have both mass and radius measured, plus equation-of-state constraints from chiral effective field theory at low density and perturbative QCD at high density, Chen et al. constrain the binary's tidal deformability to kappa2^T = 78 (+17/-11) at 68.3% credibility - a value that favors strong post-merger gravitational-wave radiation, with an estimated post-merger energy release of E_GW,p ~ 0.051 (+0.022/-0.017) solar masses c2. Assuming, as the electromagnetic counterpart modeling suggests, that the merger remnant was a supramassive neutron star, they translate this into the maximum mass of a nonrotating neutron star (M_TOV): 2.09 (+0.11/-0.09) solar masses if the remnant survived 0.1 s before collapsing, rising to 2.18 (+0.10/-0.09) solar masses for a 1 s remnant lifetime. Both values sit close to an independently reconstructed equation-of-state bound of M_TOV = 2.16 (+0.10/-0.07) solar masses, particularly once very massive neutron stars with only indirect mass measurements are excluded from that comparison sample.
M_TOV sets the boundary between the heaviest neutron star that can exist and the lightest black hole that can't be anything else, and it is one of the few numbers that pins down the neutron-star equation of state at the densities reached just before collapse - denser than anything reproducible in a terrestrial experiment. What makes this update worth recording is not the central value moving much, but that it now comes from two largely independent routes - gravitational-wave tidal deformability from one real merger, and equation-of-state reconstruction from the growing electromagnetic mass-radius catalog - landing on essentially the same number. Agreement between independent methods is what turns a model-dependent estimate into something closer to a measurement.
2
[2026-09-17] · src twelve-keys/src/exercises/rhythm_changes/exercise.py
Piano takes the whole block today: three days is the longest real gap of the three instruments right now, both alto and trumpet went in the last two days, and piano is the one that rewards a longer session rather than a quick one. It opens progressions family's third lap with the family's remaining "form" approach: not one more three-chord cadence substitute — minor_ii_v_i and backdoor both are that, and coltrane_changes already covers the family's chromatic extreme — but the eight-bar I-vi-ii-V cycle that underlies most of the standard bebop repertoire, played twice through in one chorus. Key is E♭ major, the fourth up from last session's B♭ and the first time this rotation has used E♭'s major form rather than its relative minor.
5 minutes: play the four shells alone, hands separate, very slowly, in order — Ebmaj7 (G under D), Cm7 (B♭ under E♭), Fm7 (A♭ under E♭), Bb7 (A♭ under D). Before each change, say out loud which note is about to hold and which is about to move: D holds into Cm7 as E♭ forms under it, E♭ holds again into Fm7, then D returns for Bb7 while the lower voice does the larger work of dropping from A♭ to G on the way back to the top. Naming it before playing it is the point.
The full 8 bars — I-vi-ii-V played twice — as Charleston stabs (rest, chord, rest, chord) per bar, quarter = 96, three consecutive passes with no stop between them. Success criterion: every shell lands squarely on beats 2 and 4 with no groping for the next voicing, and the top voice's holds and half-step motion (D, E♭, E♭, D) stay audibly distinct from the bottom voice's leaps — if a hand pauses to find a shell, that pass doesn't count toward the three.
Hum or sing the head of "Oleo" over your own comping, at half the tune's actual tempo since this drill only covers the A section's skeleton (the real tune runs two chords per bar and detours to IV in bars 5–6 — that's a later exercise). The goal is hearing Rollins's and Davis's phrases sit on top of exactly this cycle, not just running the changes in isolation.
Stop the moment a shell is being found by conscious search rather than landed on by habit. Three clean passes is the target, not endurance — a fourth pass spent hunting for the next voicing rehearses hesitation, not the cycle.
[2026-09-17] · src —
hearMiles Davis & the Modern Jazz Giants — Oleo (Bags' Groove) · Tidal · Spotify
Sonny Rollins built "Oleo" straight on the I Got Rhythm changes — the same I-vi-ii-V cycle today's technical target is drilling in E♭ — and this 1954 Prestige session, with Horace Silver comping the form under Rollins and Davis, is the recording rhythm_changes's own practice notes point to for singing a head over the cycle. Listen for how little the cycle needs to change to carry a whole chorus of improvising: the same four chords, twice through, under two very different solos.