{"id":"a681cc72-1d8f-408b-a1d3-c8cc87e5408b","arxiv_id":"2505.03658","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"The frequency jump in a mode-locked quantum cascade laser acts as a temporal interface supporting a non-Hermitian skin mode, observed as a sub-picosecond pulse.","lead":"Inside a fast-gain quantum cascade laser, the laser's own nonlinear steady state jumps in frequency once per round trip. The authors show this jump is a topological boundary in time that traps light fluctuations into an ultrashort 583 femtosecond pulse, and they measure and tune it.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The topological-interface mapping relies on an unquantified linear-phase approximation and on neglecting the phase-dependent potential V(F0); the actual chirped extendon needs a direct check.","rationale":"The reader's verdict and my read agree: the topological interpretation is plausible and well-supported by the measurement (spike at the frequency jump, spectral-filter chirp sweep, noise-funneling statistics) and by the full mean-field simulation. The weakest load-bearing step is the reduction of the actual chirped extendon to the constant-q0 model of Eq. (1). It is precisely this reduction that makes the spectral winding number argument and the 'topological interface' language valid; if it is only locally approximate, the claims of protection and of skin-mode confinement need qualification. The concern is not that the approximation is obviously wrong—over a 583 fs window a ~1 THz/90 ps chirp changes the local frequency by only ~6 GHz, small compared with the ~1 THz jump—but that the paper never states this estimate, never verifies W(E) for the true phase profile, and never checks the dropped V(F0) term, which has a kink by the paper's own construction. A single numerical diagonalization of the full linearized operator settles the matter. The sub-Fourier FWHM comparison is a separate weakness (583±16 vs 588 fs is within error), but it is not the central claim, so I do not hinge the verdict on it. No change to the reader's conditional verdict is needed; the requested check should be part of the revision.","tokens_in":18873,"tokens_out":15299,"duration_ms":164222,"concrete_test":"Take the numerically converged extendon F0 from the mean-field model in Eq. (D11). Linearize around it without imposing a linear-phase ansatz, keeping the V(F0) term in Eq. (B6), and solve for the eigenmodes of the Bogoliubov operator; then locate any mode localized at the phase kink and compare its FWHM with the measured 583±16 fs. Repeat the same eigenmode calculation with V(F0) artificially set to zero and with φ replaced by its local tangent ±|q0|η. If the localized mode persists in all three cases with essentially the same width, the topological mapping is robust; if it disappears or changes significantly when V(F0) or phase curvature is restored, the central claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that the extendon frequency jump is a topological NHSE interface—rests on the step from Eq. (3) to Eq. (1) via Appendix C. That mapping assumes (i) the phase is exactly linear on each side of the jump, dφ/dη = ±|q0|, and (ii) the potential V(F0) in Eq. (B6) is negligible. The actual extendon is a monotonic chirp (Figs. 2d, 3d), so φ is quadratic over the bulk and only locally approximates constant slope. The paper does not quantify the residual phase curvature over the ~583 fs skin-mode decay length, nor does it show that the spectral winding number W(E) of Eq. (A3) remains ±1 for a position-dependent q(η)=dφ/dη. Additionally, Section 2.1 states V(F)∝φ|F|^2; since the steady state F0 has a phase kink, V(F0) is kinked as well, not slowly varying as assumed in Appendix B. A kinked scalar potential at the interface could localize modes for non-topological reasons, or could shift the local dispersion enough to alter the winding-number reversal. The measured spike and noise funneling are consistent with the topological picture but do not by themselves exclude these alternative mechanisms.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the experimental observation and theoretical interpretation of an ultrafast non-Hermitian skin mode localized at the abrupt jump in instantaneous frequency of an 'extendon' state in a fast-gain quantum cascade laser. The authors derive a Bogoliubov fluctuation equation, map it to a non-Hermitian topological model under a linear-phase assumption, and identify the frequency jump as a topological interface between two bulks with opposite spectral winding numbers. Using asynchronous upconversion sampling (ASUPS), they measure an intensity spike with FWHM 583 ± 16 fs at the frequency jump, demonstrate tunability with RF modulation, and show noise funneling to the interface in both experiment and simulation. The central claim is that this is the first realization of an ultrafast topological non-Hermitian skin effect in the time domain.","tokens_in":19114,"tokens_out":4945,"duration_ms":51710,"significance":"If the central claim holds, the work is significant: it connects nonlinear laser physics to non-Hermitian topological phenomena in the time domain, provides a direct time-domain measurement of a skin mode at a sub-picosecond scale, and introduces the extendon frequency jump as a new platform for topological states of light. The paper is strengthened by the direct ASUPS measurement, the independent experimental identification of the skin mode location (not fitted), the supporting noise-funneling statistics, and the predictive numerical simulations based on a mean-field theory. The derivation in Appendices B and C is straightforward and the mapping to Eq. (1) is plausible under the stated assumptions. However, the theoretical identification depends on two assumptions—exact linearity of the phase on each side of the jump and neglect of the phase-dependent potential V(F0)—that are not quantitatively justified in the present manuscript, leaving the topological interpretation not fully established.","major_comments":[{"comment":"The mapping from the fluctuation equation (3) to the topological model (1) assumes exactly linear phase, dφ/dη = ±|q0| on each side of the frequency jump, as stated in §2.1 and used in Eq. (C8). However, the extendon is described as a monotonic chirp (Figs. 2d and 3d), meaning the phase is quadratic over the bulk and the slope is only locally constant. The manuscript does not quantify the residual phase curvature over the ~583 fs decay length of the skin mode, nor does it show that the spectral winding number W(E) of Eq. (A3) remains ±1 when q is position-dependent. I request either a quantitative estimate of the error introduced by the linear-phase approximation or a direct numerical solution of Eq. (3) with the actual extendon phase to confirm that the localized mode at the interface is the non-Hermitian skin mode.","section":"§2.1 and Appendix C"},{"comment":"The derivation of the fluctuation equation neglects the potential V(F0) on the assumption that it is 'much slower than the abrupt jump'. This assumption is questionable because, as stated in §2.1, V(F) ∝ φ|F|^2; the phase kink in the steady state F0 therefore produces a kink in V(F0) at exactly the interface. A kinked scalar potential can itself support bound fluctuations for non-topological reasons, and it can also locally shift the dispersion enough to alter the winding-number reversal that is central to the topological argument. The paper should estimate the magnitude of V(F0) relative to Γ and D and demonstrate, for example by solving Eq. (B6) with and without V(F0), that the observed and simulated mode is indeed the topological skin mode rather than a potential-bound state.","section":"Appendix B, Eq. (B6)"},{"comment":"The measured intensity spike at the frequency jump is the key experimental evidence, but the current analysis does not exclude non-topological explanations such as an intensity anomaly generated directly by the chirp discontinuity or by the phase-kinked potential discussed above. The noise-funneling data in Fig. 5 and the tunability in Fig. 4 are supportive but qualitative. I ask for a control test—for example, modifying the sign pattern of the chirp or introducing a phase kink without reversing the winding—to show that the localization is tied specifically to the winding-number reversal rather than to the interface's local potential or singular chirp.","section":"§2.2 and Fig. 3"}],"minor_comments":[{"comment":"In the fourth paragraph, 'novel meteorological and communication applications' should be 'metrological and communication applications'.","section":"Discussion"},{"comment":"The measured FWHM of 583 ± 16 fs overlaps with the transform-limited value of 588 fs within the uncertainty; the claim that the measured mode is 'narrower than the bandwidth limit' would benefit from a statement of the statistical significance or a clearer comparison that accounts for the 100 fs probe convolution.","section":"Fig. 4a"},{"comment":"The description of the optical filter measurement (Fig. 3d) is brief; it would be helpful to explain explicitly how a spectral filter combined with ASUPS yields the instantaneous-frequency map and how the frequency jump is extracted from those data.","section":"§2.2"},{"comment":"The definition of the spectral winding number in Eq. (A3) is terse; specifying the integration contour, the base point E, and the assumptions on the analyticity of det(D_q'(q')−E) would make the computation self-contained and easier to verify.","section":"Appendix A"},{"comment":"In Eq. (D11) and the paragraph after it, the average ⟨K⟩ is used before it is defined; reorder the definitions or add a pointer so that the notation is unambiguous.","section":"Appendix D"}],"recommendation":"major_revision","confidential_remarks":"The theoretical framework leans heavily on the authors' companion preprint (Ref. [36]); the experimental measurement is independent, which is good, but the manuscript should make the mapping from the physical extendon to that model airtight. If the authors can quantify the two key assumptions (linear phase and negligible V(F0)) with a direct numerical check, the paper would be considerably stronger. I also suggest the editor ensure the companion preprint is available or accepted, as the present paper's central topological argument relies on it."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know about this paper: it contains a genuinely new measurement—a sub-picosecond intensity spike in a quantum cascade laser that sits exactly at the jump in instantaneous frequency of the extendon state, can be shifted with RF detuning and power, and is accompanied by noise funneling into the same location. That is a real result, and the ASUPS technique they use to see it is impressive. The filter sweep in Fig. 3d is the most convincing piece: it directly shows the chirp discontinuity and the spike anchored there.\n\nThe theory side is plausible but softer than the abstract suggests. The mapping of the fluctuation equation to the continuum NHSE model of Ref. [36] assumes the steady-state phase is exactly linear on each side of the jump, dφ/dη = ±|q0|, with a constant slope over the whole bulk. The actual extendon has a monotonic chirp, so the phase is quadratic over most of the cycle; the winding-number argument as written applies only in a locally-linear idealization. The stress-test note is right that this is unquantified: they never check the residual phase curvature over the 583 fs decay length, nor compute W(E) for a position-dependent slope. Also, V(F0) is phase-dependent (V ∝ φ|F|²) and the kink in φ is exactly what makes the jump; dropping it in Appendix B because it is “slow” is not obviously legitimate at the interface. A kinked potential can localize modes without any topology, so the measured spike, while consistent with the NHSE picture, does not alone prove it.\n\nThe sub-Fourier-limit claim is the weakest spot. Measured FWHM 583 ± 16 fs; the transform-limited pulse from the same spectrum is 588 fs. Those overlap within the error bar. The paper says “narrower than both transform-limited pulses” but that only holds if you ignore the uncertainty. The comparison to the 619 fs convolved pulse is apples-to-oranges for the bandwidth limit. This should be fixed with a proper uncertainty propagation before anyone cites the sub-Fourier claim.\n\nMinor issues: the SI is not on arXiv, no data/code deposit, and the model in Eq. (1) comes from the coauthors' companion preprint—that is fine given Ref. [36] exists, but it raises the circularity burden to a level the paper does not fully address. I would not call this a fatal flaw; the measurement is independent and the mapping is plausible.\n\nWho gets value: people working on time-varying topological photonics and mode-locked lasers. It deserves a serious referee, but I would send it back for major revision: quantify the phase-curvature and potential-kink effects, tighten the Fourier-limit statistics, and release the data or at least the SI. Under that path, I would cite it as the first direct time-domain NHSE observation in a laser.","headline":"A real experimental observation of a sub-ps spike locked to the extendon's frequency jump, but the topological-interface interpretation is built on approximations that need quantitative checking, and the sub-Fourier-limit claim does not survive the error bar.","tokens_in":19710,"tokens_out":1571,"would_cite":true,"duration_ms":17921,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that the abrupt jump in instantaneous frequency of a laser's nonlinear extendon state acts as a topological interface that binds non-Hermitian skin modes of the field fluctuations, measured as a 583 ± 16 fs intensity…","keywords":["non-Hermitian topology","skin effect","spatiotemporal topology","quantum cascade lasers","extendon","Bogoliubov fluctuations","asynchronous upconversion sampling","ultrafast mode-locking"],"falsifier":"Use filter-resolved asynchronous sampling to reconstruct $\\phi(\\eta)$ over a full half-cycle and compute the phase residual $\\phi(\\eta) - q_0\\eta$ in the bulk. If away from the jump this residual deviates by order a radian or more (equivalently, $d\\phi/d\\eta$ varies by order $|q_0|$ over the half-cycle), the two sides are not topological bulks with winding $\\pm 1$, and the predicted skin mode should either disappear or detach from the jump; observing the spike under such conditions would falsify the mapping.","tokens_in":18652,"feed_emoji":"⚡","tokens_out":7843,"duration_ms":77787,"temperature":0.7,"pith_summary":"Nonlinear mode-locking in a fast-gain semiconductor laser produces an extendon: a constant-intensity field whose phase sweeps linearly in time and then resets, so the instantaneous frequency jumps once per round trip. This paper claims that the phase jump is a topological interface for small fluctuations riding on the extendon. Because the phase slope is opposite on the two sides of the jump, linearized fluctuations live in non-Hermitian bulk phases with opposite spectral winding numbers, and the fluctuation dynamics maps exactly onto a momentum-shifted parametric model whose skin modes localize at the sign change. The authors verify this in a quantum cascade laser by asynchronous sampling of the output: they see a 583 ± 16 fs intensity spike locked to the frequency jump, find that noise is funneled toward that same point, and show the pulse's width and arrival time can be steered with RF modulation. If the identification is right, it gives a spatial-lattice-free route to topologically bound, sub-transform-limited pulses of light in time.","feed_headline":"A frequency jump in a laser traps light in 583-fs skin modes","feed_subtitle":"Field fluctuations in a quantum cascade laser pile into a topologically bound pulse locked to the frequency jump.","key_machinery":"The load-bearing object is the extendon's phase kink combined with the Bogoliubov coupling generated by the saturated-gain term. The paper treats the phase as exactly linear on each side, $\\phi(\\eta) = \\mp q_0\\eta$, so the exponential coupling $e^{2i\\phi(\\eta)}$ becomes a momentum shift $2q_0$ in the conjugate-pairing term. The machinery then is the spectral winding number $W(E) = \\int dq'/(2\\pi i)\\,\\partial_{q'}\\log\\det(D_{q'} - E)$ of the $2\\times 2$ dynamical matrix $D_{q'}$; it returns $+1$ on one side and $-1$ on the other, and the sign reversal at the reset point is what forces a skin mode to localize there. Experimentally, the machinery is completed by asynchronous upconversion sampling, which resolves a sub-picosecond feature without being washed out by phase noise.","core_discovery":"The paper establishes that the extendon steady state $F_0 = \\sqrt{P_0}e^{i\\phi(\\eta)}$ carries a phase profile $\\phi(\\eta)$ that is linear on each side of a reset point with $d\\phi/d\\eta = \\pm |q_0|$. Linearizing the mean-field laser equation for the fluctuation $\\delta F$ gives $i\\partial_T \\delta F = -D\\partial_\\eta^2 \\delta F - i\\Gamma\\delta F - i\\Gamma e^{2i\\phi(\\eta)}\\delta F^*$; substituting constant-slope phases and Fourier transforming reproduces, term by term, the continuum model $H = \\int dq\\,[(q^2/2m - i\\Gamma)a_q^\\dagger a_q - i\\Gamma a_q^\\dagger a_{-q-2q_0}^\\dagger + \\mathrm{h.c.}]$. That model has spectral winding numbers $W(E) = \\pm 1$ on the two sides, so a sign flip of $q_0$ is a domain wall, and the paper predicts and measures skin modes bound there: an intensity spike of width $583 \\pm 16$ fs sitting exactly at the jump in instantaneous frequency. The same funneling is seen in the noise statistics, where fluctuations accumulate at the interface.","pith_inferences":["An extension the paper leaves implicit: if the linear-slope idealization holds, the effect should be generic to any source that forms a phase kink with opposite linear slopes, not only quantum cascade lasers, so strongly chirped mode-locked lasers and microcombs are natural places to look for the same sub-picosecond skin mode.","Because the paper's own model ties the mode width to the size of the frequency jump, engineering larger dispersion or stronger phase modulation could push the confinement well below the 583 fs measured here, possibly approaching the ~100 fs probe limit.","A chain of several phase jumps per round trip would create multiple interfaces with alternating winding, offering a synthetic time-lattice of skin modes whose spacings could be tuned independently, something the paper does not simulate.","If the confinement is truly topological, a controlled experiment that linearizes the chirp only on one side of the jump should leave the mode on the other side nearly unchanged; absence of that asymmetry would indicate the localization is ordinary spectral filtering rather than skin-effect topology."],"forward_implications":["The measured skin mode is shorter than the transform-limited pulse from the same optical spectrum (583 ± 16 fs versus 588 fs and 619 fs), so the topological mechanism can generate pulses below the bandwidth limit of the underlying field.","Enlarging the frequency jump, here by increasing RF modulation power, sharpens and strengthens the spike, giving a direct control parameter for pulse duration.","Detuning the modulation frequency across the locking range moves the interface within the optical cycle in a quasi-linear way, so the arrival time of the topological pulse is predictable and adjustable; at high modulation depth the position shifts by about $\\pi/2$.","Local fluctuations launched at any intracycle time propagate toward the interface, and white-noise simulations plus intensity statistics show the excess noise accumulates at the jump, meaning the laser self-funnels energy into the topological state rather than spreading it."],"supporting_citations":[{"why":"Supplies the continuum non-Hermitian model, the spectral winding number, and the prediction of skin modes localized at a sign flip of $q_0$, which the paper maps its fluctuations onto.","marker":"[36]"},{"why":"Provides the mean-field generalized nonlinear Schrödinger equation that gives the extendon steady state and underlies the fluctuation expansion.","marker":"[76]"},{"why":"Introduces the asynchronous upconversion sampling method used to measure the sub-picosecond intensity spike and its position relative to the frequency jump.","marker":"[78]"},{"why":"Gives the fast-gain suppression argument and the predicted timing shift of the extendon used to explain the skin-mode position tuning.","marker":"[79]"},{"why":"Supplies the fluctuation formalism used to write the linearized Bogoliubov equation for $\\delta F$.","marker":"[82]"},{"why":"Establishes the microwave injection locking scheme that stabilizes the repetition rate and lets the authors tune the frequency jump via bias modulation.","marker":"[85]"}],"fun_headline_variants":["Light trapped in 583-fs skin modes at a frequency jump","Ultrafast skin effect: laser's frequency jump binds light","Topological light modes pinned by a laser frequency jump","583-fs skin modes from a laser's non-Hermitian jump","Light funneled into a 583-fs topological interface"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes the extendon phase is exactly linear in the co-rotating coordinate over the whole region on each side of the jump, $d\\phi/d\\eta = \\pm |q_0|$, and that the slow nonlinear potential $V(F_0)$ changes too gently to matter; if the chirp is curved instead of constant, both the winding-number assignment and the skin-mode confinement lose their exact footing.","fun_headline_variants_meta":{"raw":{"variants":["Light trapped in 583-fs skin modes at a frequency jump","Ultrafast skin effect: laser's frequency jump binds light","Topological light modes pinned by a laser frequency jump","583-fs skin modes from a laser's non-Hermitian jump","Light funneled into a 583-fs topological interface"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00053,"raw_usage":{"total_tokens":2594,"prompt_tokens":1025,"completion_tokens":1569,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":641,"completion_tokens_details":{"reasoning_tokens":1483}},"tokens_in":641,"tokens_out":1569,"duration_ms":11001,"temperature":1.0,"reasoning_tokens":1483,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T23:45:35.625577+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Use filter-resolved asynchronous sampling to reconstruct $\\phi(\\eta)$ over a full half-cycle and compute the phase residual $\\phi(\\eta) - q_0\\eta$ in the bulk. If away from the jump this residual deviates by order a radian or more (equivalently, $d\\phi/d\\eta$ varies by order $|q_0|$ over the half-cycle), the two sides are not topological bulks with winding $\\pm 1$, and the predicted skin mode should either disappear or detach from the jump; observing the spike under such conditions would falsify the mapping.","supporting_citations":[{"cited_title":"Optica 7(12), 1781 (2020) https://doi.org/10.1364/ optica.408917","cited_arxiv_id":null,"evidence_quote":"Provides the mean-field generalized nonlinear Schrödinger equation that gives the extendon steady state and underlies the fluctuation expansion."},{"cited_title":"Nanophotonics (2025) https://doi.org/10.1515/ nanoph-2024-0768","cited_arxiv_id":null,"evidence_quote":"Gives the fast-gain suppression argument and the predicted timing shift of the extendon used to explain the skin-mode position tuning."},{"cited_title":"Europhysics Letters 102(6), 67007 (2013) https://doi.org/10.1209/ 22 0295-5075/102/67007","cited_arxiv_id":null,"evidence_quote":"Supplies the fluctuation formalism used to write the linearized Bogoliubov equation for $\\delta F$."},{"cited_title":"Laser & Photonics Reviews 15(12), 2100242 (2021) https: //doi.org/10.1002/lpor.202100242","cited_arxiv_id":null,"evidence_quote":"Establishes the microwave injection locking scheme that stabilizes the repetition rate and lets the authors tune the frequency jump via bias modulation."}],"review_version":1}