{"id":"93c3a5f0-ab19-49d3-93bc-87da1ce60a54","arxiv_id":"2509.01950","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Space-time wave packets can be designed to follow arbitrary power-law curved trajectories with symmetric profiles and scale-independent acceleration.","lead":"Researchers shaped light pulses called space-time wave packets so their center follows curved paths, including quadratic, cubic, and square-root power laws. These 'bending' beams keep symmetric profiles and offer a new way to steer light around obstacles.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The bending-trajectory design relies on an approximate, underived SLM-position-to-axial-distance mapping and a phase encoding that may not impart the intended local spatial frequency; no quantitative trajectory fit is provided.","rationale":"The reader identified the underived xs/z mapping as the weakest assumption; I agree that this is the critical step connecting the SLM phase to the curved trajectory. However, I sharpen it with an additional internal-consistency concern: the phase expression as written would not generally produce the intended local spatial frequency, since the local frequency is the derivative of the phase. This makes the algorithm under-specified even if the mapping were correct. I also note that the lack of quantitative trajectory extraction weakens the experimental support. These concerns do not refute the central claim; they motivate the same conditional verdict: the authors should provide a derivation or validation of the mapping, confirm the phase-encoding relation, and release quantitative trajectory fits. Hence the reader's CONDITIONAL verdict is appropriate and unchanged.","tokens_in":7632,"tokens_out":4823,"duration_ms":60550,"concrete_test":"Digitize the measured I(x,z) in Fig. 4(c-e), extract the peak position as a function of z, and fit to x_o(z)=A z^gamma with gamma free; report residuals and compare gamma to the target (2, 3, 0.5). Additionally, compute the Fourier transform of the nominal SLM phase Phi(omega,x_s)=k'_x(omega,x_s)x_s and compare the resulting local spatial-frequency distribution with the designed k'_x(omega,z(x_s)). If the peak of the distribution differs from k'_x by more than the spectral resolution, the phase encoding must be corrected to Phi=integral k'_x dx_s and the experiment re-analyzed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that measured peaks follow x_o(z)=x_1(z/z_1)^gamma depends on the algorithm's mapping from SLM coordinate x_s to axial distance z. The paper states x_s/z ~ k'_x/k'_z, cites refs [26,27], and does not derive it; this is especially insecure for large transverse displacements (x_1 up to 200 um over z_1=40 mm) where paraxial/geometric approximations may break. More concretely, after replacing z by x_s, the displayed phase is written as Phi(omega,x_s)=k'_x(omega,x_s)x_s. The local spatial frequency imparted at x_s is dPhi/dx_s = k'_x + x_s dk'_x/dx_s, not k'_x. Thus the actual k_x content of the synthesized wave packet may differ from the designed z-dependent tilt unless k'_x is nearly independent of x_s (which is not the case for quadratic, cubic, or square-root trajectories) or the phase was implicitly integrated. Without a derivation or a direct spectral measurement verifying that each x_s carries the intended k'_x, the realized trajectory cannot be inferred from the phase pattern. The paper also provides no extracted peak positions, fits, or residuals for Fig. 4(c-e), so the experimental realization is only qualitative. This is load-bearing because the claimed arbitrary power-law trajectories are generated by this mapping and encoding step.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the experimental realization of 'bending space-time wave packets' (STWPs): pulsed beams with symmetric transverse profiles whose time-averaged intensity peak travels along curved trajectories of the form x_o(z)=x_1(z/z_1)^γ, with γ = 1, 2, 3, and 1/2 demonstrated. The design algorithm starts from a propagation-invariant STWP and rotates its spatiotemporal spectral support by a z-dependent angle, using the approximate mapping x_s/z ≈ k'_x/k'_z to convert axial position into the SLM coordinate x_s. The SLM phase is then set to Φ(ω,x_s)=k'_x(ω,x_s)x_s. Measured spatiotemporal spectra and time-averaged intensity profiles are presented for each power-law case. The paper claims that this approach yields self-accelerating beams with symmetric profiles and acceleration rates independent of the beam spatial scale.","tokens_in":7969,"tokens_out":7449,"duration_ms":82802,"significance":"If fully validated, the result would establish a new class of self-accelerating optical beams with three distinctive features: symmetric transverse profiles, arbitrary positive power-law trajectories (including fractional exponents), and acceleration controlled independently of the spatial scale of the beam. This goes beyond Airy beams and would be of interest for applications such as target avoidance and for fundamental studies of spatiotemporal wave-packet propagation. The experimental implementation is built on a mature STWP synthesis platform and the visual evidence for bending in several power-law cases is striking. However, the current manuscript lacks the quantitative trajectory analysis needed to substantiate the central claim, and the phase-encoding step in the design algorithm is described in a way that is internally questionable.","major_comments":[{"comment":"The displayed phase is written as Φ(ω,x_s)=k'_x(ω,x_s)x_s. The local spatial frequency imparted by the SLM is ∂Φ/∂x_s = k'_x + x_s ∂k'_x/∂x_s, not k'_x. For the power-law trajectories used here, k'_x varies with x_s through the mapping x_s/z≈k'_x/k'_z. For example, for γ=2, k'_x∝x_s^{1/2} and ∂Φ/∂x_s=(3/2)k'_x; for γ=1/2, k'_x∝x_s^{-1} so Φ is constant across x_s and no z-dependent tilt is encoded. Unless the phase was actually computed as Φ=∫k'_x dx_s, or direct spectral measurements show that each x_s carries the intended k'_x, the realized k_x content cannot be inferred from the phase pattern. This is load-bearing because the claimed trajectories are generated by this step.","section":"Algorithm for designing the spatiotemporal spectral phase"},{"comment":"No extracted peak positions, trajectory fits, or residuals are provided for Fig. 4(c–e). The reader is shown target I(x,z) next to measured I(x,z), but the claimed quantitative agreement with x_o(z)=x_1(z/z_1)^γ is not established. Because the trajectory is directly encoded into the SLM phase, the observation of bending is not an independent test of a prediction; the scientific content is the transfer function from designed phase to realized trajectory. Please plot the measured x_o(z) for each case, overlay the target curve, and report fit parameters with uncertainties.","section":"Measurement results"},{"comment":"The mapping x_s/z≈k'_x/k'_z is cited to Refs. [26,27] but not derived or validated here. In the algorithm, k'_x(ω,z) appears on both sides of this mapping, so it is unclear how z is eliminated in favor of x_s without a self-consistent solution. The accuracy of the mapping is especially questionable for the large transverse displacements (x_1 up to 200 µm over z_1=40 mm) and for γ=1/2, where the local tilt angle diverges as z→0. Provide a derivation, a numerical test of the mapping against the designed k'_x, or a spectral measurement that directly verifies the encoded k'_x at each x_s.","section":"Algorithm for designing the spatiotemporal spectral phase"},{"comment":"The abstract and conclusion claim that the acceleration rate is independent of the beam spatial scale, but no experiment varies the transverse profile scale while holding the trajectory fixed. The statement that 'these power laws are all associated with the same transverse profile shape and scale' does not demonstrate independence. Either add a comparison at different spatial scales or temper the claim to what is actually shown.","section":"Discussion"}],"minor_comments":[{"comment":"The text for Fig. 4(d) says x_1=200 µm, while the caption says x_1=150 µm. Also, 'x_1(z)=x_1(z/z_1)^3' should read 'x_o(z)=x_1(z/z_1)^3'.","section":"Fig. 4"},{"comment":"The expression for the propagation-invariant STWP phase is written as Φ(ω,x_s)=k_x(ω)x, but the spatial variable should be x_s for consistency.","section":"Algorithm"},{"comment":"The third row is labeled 'target time-averaged intensity profile I(x,z)' but the method used to compute it from the designed spectrum is not stated. Please specify how the target intensity is calculated.","section":"Fig. 4"},{"comment":"The symbol x_o is used for the profile center, but in the paragraph on tilted STWPs the trajectory is written as x_o(z)=z tan ϕ_o; later the text uses x_1(z) for the cubic trajectory. Please unify notation.","section":"Notation"}],"recommendation":"major_revision","confidential_remarks":"The phase-encoding issue is the main technical hurdle: the literal formula Φ=k'_x x_s does not produce the intended local spatial frequency when k'_x depends on x_s, and for γ=1/2 it predicts no bending. If the authors can clarify that the phase was integrated or otherwise correct the description, and if they add quantitative trajectory fits, the central claim would become credible. The lack of public data is not a block, but the absence of quantitative comparisons is."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is worth a look: it demonstrates experimentally that space-time wave packets can be made to follow curved power-law trajectories, including the first square-root case, while keeping a symmetric transverse profile. That is genuinely new and, if it holds, useful for structured-light and target-avoidance work. The authors give a clear algorithmic procedure for designing the SLM phase, and the measured spatiotemporal spectra and intensity images are consistent with the intended behavior. I believe the central idea is sound and the experiments are not faked.\n\nThe soft spots are real, though. The design step maps SLM coordinate to axial distance via x_s/z ~ k'_x/k'_z and then writes the phase as Phi(omega,x_s)=k'_x(omega,x_s)x_s. The local spatial frequency imparted by that phase is dPhi/dx_s = k'_x + x_s dk'_x/dx_s, not k'_x. Unless k'_x is effectively independent of x_s over the relevant range—which is not the case for the quadratic, cubic, or square-root trajectories—the actual k_x content differs from what the algorithm assumes. The paper does not derive the mapping or validate it with a direct spectral measurement. This is a load-bearing step, because the claimed arbitrary trajectories are generated by exactly this encoding. The fix is probably to integrate the phase (Phi = integral k'_x dx_s), but as written the algorithm is incomplete at best.\n\nSecond, the experimental verification is qualitative. Figure 4 shows the target curves drawn over the measured intensity, but there are no extracted peak positions, fits, or residuals. For a paper whose central claim is that the beam follows x_o(z)=x_1(z/z_1)^gamma, that is a surprising omission. The data availability statement also says no public data, which makes independent verification harder.\n\nI do not think these problems are fatal. The experimental evidence is visually convincing, and the underlying concept—building on Liang et al. and the earlier axial-encoding work—is reasonable. The paper should be sent to referees, but the authors need to either derive the mapping properly, correct the phase encoding, or show with measured spectra that each SLM position carries the intended k'_x. They should also include quantitative trajectory fits. If they can do that, this becomes a solid contribution.\n\nReading group: yes, for the discussion of how to design bending STWPs. I would not cite it in its current form because of the phase-encoding issue, but I would cite a corrected version.\n\nRecommendation: send to peer review with major revision requirements, not desk reject.","headline":"A clever experimental demonstration of bending STWPs with power-law trajectories, but the design algorithm's phase encoding has a derivative inconsistency and the trajectory fits are only qualitative.","tokens_in":8408,"tokens_out":3670,"would_cite":false,"duration_ms":40962,"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":"Sculpting a pulse's spatiotemporal spectrum lets a symmetric beam's peak follow linear, quadratic, cubic, or square-root curves without diffraction.","keywords":["space-time wave packets","bending STWPs","self-accelerating beams","spatiotemporal spectrum","power-law trajectories","angular dispersion","diffraction-free propagation","Airy beams"],"falsifier":"Encode a trajectory with a steep exponent or large total displacement, then measure the time-averaged intensity I(x,z) across the full designed range and fit the peak. If the measured peak deviates from x1(z/z1)^γ by more than the beam's transverse width near the end of the range—or if the (kx,λ) spectral projection fails to show the predicted finite-bandwidth spread at each wavelength—the xs≈z approximation is the point of failure.","tokens_in":7532,"feed_emoji":"🌀","tokens_out":7237,"duration_ms":77308,"temperature":0.7,"pith_summary":"The paper claims that by deviating from the one-wavelength-to-one-spatial-frequency rule that defines standard space-time wave packets, an optical pulse can be made to accelerate sideways along a prescribed power law. The resulting 'bending STWPs' keep a symmetric, diffraction-free transverse profile, so the curve of the trajectory is no longer tied to an asymmetric beam shape as it is for Airy beams. The authors demonstrate this experimentally for linear, quadratic, cubic, and square-root trajectories and report that the acceleration rate does not depend on the beam's spatial scale. This matters because it separates trajectory design from beam-profile design and suggests a route to steering light around line-of-sight obstacles.","feed_headline":"Symmetric light beams follow curved paths of any power law","feed_subtitle":"Shaping a pulse's spatiotemporal spectrum sends its peak along any power-law curve, without diffraction.","key_machinery":"The load-bearing object is the bending STWP's spectral support: instead of the intersection of the free-space light cone with a single tilted plane (a one-dimensional curve that enforces one spatial frequency per wavelength), the bending design uses a two-dimensional domain swept out by continuously rotating that tilted plane around the ω/c axis. The algorithm realizes this domain by assigning each wavelength a finite-bandwidth spatial spectrum k'_x(ω,z) that varies with axial position z, and uses the map xs/z ≈ k'_x/k'_z to place those frequencies on the SLM. The phase Φ(ω,xs) = k'_x(ω,xs)xs then encodes the trajectory. The same profile shape and scale are reused for every exponent, demonst","core_discovery":"The central discovery is an algorithmic method for choosing the two-dimensional phase pattern on a spatial light modulator so that a space-time wave packet's spectral support on the light cone becomes a two-dimensional domain rather than a curve. Each wavelength is paired with a finite range of spatial frequencies, and each spatial-frequency component is placed at an SLM position that maps to an axial propagation distance through the approximate relation xs/z ≈ k'_x/k'_z. This effectively rotates the STWP's tilt direction continuously along z, making the time-averaged-intensity peak follow xo(z) = x1(z/z1)^γ for any positive exponent γ. Experiments with γ = 2, 3, and 0.5 show the intended cu","pith_inferences":["The xs/z ≈ k'_x/k'_z mapping suggests the algorithm should work for any monotonically increasing trajectory, not just power laws, by substituting an arbitrary xo(z); this extension is implicit but not tested in the paper.","The finite spectral bandwidth available per wavelength and the SLM pixel pitch impose a practical ceiling on how sharply a trajectory can bend, so a systematic error study across exponents and displacements would reveal where the mapping approximation breaks.","Since the transverse profile is preserved while the trajectory bends, spectral phase shaping could be layered on top to deliver a designed mode along a curved path—an experiment the paper's setup is already equipped to attempt."],"forward_implications":["Self-accelerating beams no longer need an asymmetric profile; any symmetric, diffraction-free STWP profile can be bent, and the direction of curvature is set by the spectral tilt rather than by the beam shape.","The trajectory exponent γ can be chosen independently of beam scale and profile, so linear, quadratic, cubic, and fractional power laws are all reachable from the same apparatus by changing only the SLM phase.","Because the acceleration rate is decoupled from the transverse scale, one can shrink or expand the beam without changing how quickly it bends.","The method extends the STWP toolkit to combine axial acceleration and transverse bending, which the authors identify as a route to spatiotemporal self-acceleration.","Bending STWPs make line-of-sight target avoidance directly testable: a beam with a symmetric profile can be steered around an obstacle while preserving its transverse structure."],"supporting_citations":[{"why":"Supplies the axial mapping xs/z ≈ k'_x/k'_z that the bending-phase algorithm uses to turn SLM positions into propagation distances.","marker":"[26]"},{"why":"Extends the same axial spectral-encoding relation to long distances and supports using the mapping over the experimental range.","marker":"[27]"},{"why":"Defines propagation-invariant STWPs and their light-cone spectral support, which the bending design modifies.","marker":"[13]"},{"why":"Provides the universal angular-dispersion synthesizer that produces the spatiotemporal spectra used in the experiments.","marker":"[29]"},{"why":"Supplies the synthesis and characterization details for space-time light sheets used in the measurements.","marker":"[30]"},{"why":"Introduces finite-energy Airy beams, the baseline self-accelerating beam whose asymmetric-profile constraint bending STWPs overcome.","marker":"[1]"},{"why":"Provides the experimental observation of accelerating Airy beams, the standard comparison for curved beam trajectories.","marker":"[2]"}],"fun_headline_variants":["Bending light: symmetric pulses follow any power-law curve","Space-time wave packets bend along arbitrary power laws","Curved light paths: any exponent, no diffraction","Sculpted spectra send pulses along power-law curves","Light's peak rides power-law trajectories, symmetric and tight"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The whole design hinges on the approximate mapping between a spatial-light-modulator coordinate and an axial propagation distance, xs/z ≈ k'_x/k'_z; if that mapping drifts for large transverse displacement or steeply curved trajectories, the measured peak will not follow the intended power law.","fun_headline_variants_meta":{"raw":{"variants":["Bending light: symmetric pulses follow any power-law curve","Space-time wave packets bend along arbitrary power laws","Curved light paths: any exponent, no diffraction","Sculpted spectra send pulses along power-law curves","Light's peak rides power-law trajectories, symmetric and tight"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00024,"raw_usage":{"total_tokens":1340,"prompt_tokens":713,"completion_tokens":627,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":457,"completion_tokens_details":{"reasoning_tokens":549}},"tokens_in":457,"tokens_out":627,"duration_ms":7352,"temperature":1.0,"reasoning_tokens":549,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T12:01:16.356255+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Encode a trajectory with a steep exponent or large total displacement, then measure the time-averaged intensity I(x,z) across the full designed range and fit the peak. If the measured peak deviates from x1(z/z1)^γ by more than the beam's transverse width near the end of the range—or if the (kx,λ) spectral projection fails to show the predicted finite-bandwidth spread at each wavelength—the xs≈z approximation is the point of failure.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the axial mapping xs/z ≈ k'_x/k'_z that the bending-phase algorithm uses to turn SLM positions into propagation distances."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Extends the same axial spectral-encoding relation to long distances and supports using the mapping over the experimental range."},{"cited_title":"Yessenov, L","cited_arxiv_id":null,"evidence_quote":"Defines propagation-invariant STWPs and their light-cone spectral support, which the bending design modifies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the universal angular-dispersion synthesizer that produces the spatiotemporal spectra used in the experiments."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the synthesis and characterization details for space-time light sheets used in the measurements."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces finite-energy Airy beams, the baseline self-accelerating beam whose asymmetric-profile constraint bending STWPs overcome."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the experimental observation of accelerating Airy beams, the standard comparison for curved beam trajectories."}],"review_version":1}