Pool: 246 new + 134 cross-listed (gr-qc / astro-ph.HE / hep-th, after date-based reclassification)
Candidates deep-read: 10
Survivors after critical evaluation: 7 (3 Must-Read · 2 Should-Read · 2 Worth-Skimming)
Generated: 2026-05-05
arXiv:2604.26011 — Levati, Cárdenas-Avendaño
One-line: Fisher-matrix study of how 3:2, 2:1, 3:1 and 4:3 transient resonances bias EMRI parameter recovery in LISA, with explicit treatment of signed kicks ΔE, ΔL_z, ΔQ.
Quality 7/10 · Relevance 9/10. Solid, incremental refinement of Speri & Gair (2021) with broader resonance coverage and sign-aware kick treatment. Caveat: Fisher local-Gaussian assumption is borderline at threshold SNR; mechanism for kicks rests on PN/post-adiabatic input.
Full report
arXiv:2604.25916 — Dlapa, Kälin, Liu, Porto
One-line: Closes the conservative tail tower of two-body GR scattering up to 5PM / 10SF order via worldline EFT with polylogarithms of weight three; cross-validated against six PN orders.
Quality 9/10 · Relevance 9/10. Mature, frontier calculation; SpideR sparse-IBP machinery is a useful by-product. Caveat: physical interpretation of weight-3 polylogs and phenomenological impact on EMRI waveforms is not discussed.
Full report
arXiv:2604.26253 — Zhu, Pretorius, Stone
One-line: First γ ≈ 5.1 NR encounters of equal-mass non-spinning BHs reveal a prolonged irregular-emission regime driven by transient null trapping and curvature lensing, with up to 65% of ADM energy radiated.
Quality 8/10 · Relevance 8/10. Genuine new regime + new gauge driver (telegrapher lapse). Tier 3: Pretorius. Caveats: γ = 5.1 not asymptotic; junk-radiation-subtraction procedure deferred to a companion paper; gauge-robustness untested.
Full report
arXiv:2605.00693 — Chakraborty, Saketh, Hinderer, Steinhoff
One-line: Linear-in-frequency tidal response of spinning BHs derived rigorously from BHPT with full multipole-mixing, matched to a worldline EFT action; complements the Charalambous et al. static-Love-vanishing story with non-zero dynamical Wilson coefficients.
Quality 7.5/10 · Relevance 8.5/10. Tier 2 (Steinhoff) + Tier 3 (Hinderer). Caveats: matching-scheme dependence not exhausted; lacks an independent numerical-Teukolsky cross-check.
Full report
arXiv:2604.28140 — Blanco
One-line: A clean trick — analytically continue the Darwin (p, e) parametrization through complex values to obtain a single, smooth phase variable that crosses the separatrix into the plunge.
Quality 7/10 · Relevance 7/10. Tier 2 (Blanco). Conceptually new and elegant. Caveat: Schwarzschild only, no back-reaction; the Kerr generalization is the natural next paper and would lift the relevance dramatically.
Full report
arXiv:2605.00124 — Ravichandran, Nee, Mitman, Boyle, Pfeiffer et al.
One-line: NR-trained surrogates for final mass / spin / recoil and for eccentricity & mean-anomaly evolution, q ≤ 4, e ≲ 0.23, nonspinning; ships systematic-error bars.
Quality 8/10 · Relevance 5/10. Tier 2 (Nee) + Tier 3 (Mitman, Boyle, Pfeiffer). Useful as a benchmark target for EOB/PN models; coverage too narrow for LISA EMRIs.
Full report
arXiv:2604.27429 — Sharma, Heger, Price, Tejeda, Grishin, Manzaneda, Trani
One-line: First systematic side-by-side test of seven different GR/PN/metric-with-perturbation codes for the binary-star + SMBH three-body problem; identifies a robust spurious pericentre decay in pair-wise PN.
Quality 7/10 · Relevance 6/10. Useful “which-code-to-trust” reference for galactic-centre stellar dynamics; not new physics by itself.
Full report
Two clean themes dominate the high-quality survivors. First, the high-precision two-body program continues its impressive march: Dlapa et al.’s 5PM / 10SF tail completion and the Chakraborty–Hinderer–Steinhoff dynamical-Love EFT match are part of the same machine — squeezing the conservative and tidal sectors of binary inspiral to the level where the post-adiabatic EMRI program can absorb them. The fact that the SF expansion has reached 10th order in the small-mass-ratio direction simultaneously with the PM expansion reaching 5th order means we are now firmly in the regime where the cross-checks between the two pictures are meaningful, not just consistency-trivialities.
Second, the resonance / mean-anomaly literature is maturing. Levati & Cárdenas-Avendaño on EMRI transient-resonance bias, together with the eccentricity-and-mean-anomaly surrogate work of Ravichandran-Nee-Mitman-Boyle, both reflect a community-wide shift from “do mean anomaly and resonances matter at all?” (settled: yes) to “what is the production-grade pipeline for them?” The Speri-Gair paper of 2021 may well be cited as the one-paper origin of this entire sub-field.
Outside those two trends, Zhu-Pretorius-Stone’s γ ≈ 5.1 NR letter is a stand-alone strong-field NR result with mechanism content (transient null trapping, ω⁻⁵/³ tail). Blanco’s plunge-Darwin trick is a small but real piece of analytical structure that in particular resonates with the action-angle / Hamilton-Jacobi line of work.
These are replacement (v2+) papers rather than fresh submissions, surfaced because they touch the core program or come from close collaborators.