{"id":"8004e5cb-9bed-45ae-8f2d-c4e2f89866e5","arxiv_id":"2606.30936","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Conditional normalizing flows model the posterior over initial states from short-arc angles-only measurements in cislunar NRHOs and supply warm starts for nonlinear least-squares refinement.","lead":"The paper trains a conditional normalizing flow on simulated angles-only observations from Near Rectilinear Halo Orbits to estimate the distribution of initial spacecraft states in cislunar space. A smart generalist might read it to understand how generative AI models can generate candidate solutions for orbit determination problems where traditional methods face ambiguity.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"Generalization of NRHO-trained conditional flow to other cislunar orbits or real noise not shown","rationale":"The reader's weakest assumption directly identifies the same generalization gap. Because the manuscript provides no OOD or real-data validation, the UNVERDICTED verdict is unaffected.","tokens_in":1631,"tokens_out":295,"duration_ms":14089,"concrete_test":"Generate a test set of short-arc topocentric angles from a non-NRHO cislunar orbit (e.g., a southern L2 halo or Lyapunov orbit) using the same dynamics model but different initial conditions; draw 1000 samples from the trained conditional flow, run NLS refinement on each, and report the fraction that converge to <1 km position error versus a random-initialization baseline.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central construction trains a conditional normalizing flow exclusively on simulated topocentric angles from perturbed NRHO trajectories. The claim that the resulting posterior yields statistically consistent, physics-informed hypotheses for general angles-only cislunar OD therefore rests on the untested assumption that the learned conditional density transfers outside the NRHO training distribution. No mention is made of hold-out orbit families, different dynamical regimes, or real sensor noise models; if the flow has overfit to NRHO-specific geometry and observability, the sampled states will not remain accurate or provide reliable warm starts for NLS on out-of-distribution arcs.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript proposes training a conditional normalizing flow on perturbed topocentric angles from simulated Near Rectilinear Halo Orbit (NRHO) trajectories to perform angles-only orbit determination in the cislunar regime. Given new short-arc measurements, the flow models the posterior over initial states, which is sampled to produce candidate state vectors that are then refined by nonlinear least-squares to supply warm starts for classical estimators.","tokens_in":1735,"tokens_out":386,"duration_ms":19695,"significance":"If the central claims hold, the work would demonstrate a practical use of conditional normalizing flows to represent potentially multimodal posteriors in a physics-constrained setting, offering an alternative initialization strategy for angles-only cislunar OD where traditional batch estimators can be sensitive to poor initial guesses. The explicit coupling of generative density estimation with subsequent NLS refinement is a methodological strength that could be extended to other under-observed regimes.","major_comments":[{"comment":"Abstract: the claim that the sampled states are 'statistically consistent and physics-informed' and provide 'competitive warm starts' cannot be evaluated because the abstract (and the description supplied) contains no quantitative results, error metrics, baseline comparisons, or validation details on either in-distribution or out-of-distribution cases.","section":"Abstract"},{"comment":"Abstract / training description: the method is trained exclusively on perturbed topocentric observations from NRHO trajectories, yet the central claim of applicability to 'general angles-only cislunar OD' is not supported by any reported hold-out tests on other orbit families, different dynamical regimes, or realistic sensor noise models; if the conditional density overfits to NRHO-specific geometry and observability, the generated hypotheses will not remain accurate or reliable for NLS initialization outside the training distribution.","section":"Abstract"}],"minor_comments":[],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the constructive comments on our manuscript. We agree that the abstract requires quantitative support and that the scope of validation should be clarified. We have revised the abstract accordingly and added explicit discussion of limitations. Point-by-point responses follow.","responses":[{"response":"We agree that the original abstract was too concise and omitted supporting metrics. The revised abstract now includes summary quantitative results from the experiments, such as median position and velocity errors on the test set, the fraction of samples that converge under NLS refinement, and a direct comparison against a random-initialization baseline. These additions allow the claims to be evaluated without requiring the reader to consult the full text.","revision_made":"yes","referee_comment":"[Abstract] Abstract: the claim that the sampled states are 'statistically consistent and physics-informed' and provide 'competitive warm starts' cannot be evaluated because the abstract (and the description supplied) contains no quantitative results, error metrics, baseline comparisons, or validation details on either in-distribution or out-of-distribution cases."},{"response":"The work deliberately targets NRHOs because they represent a demanding cislunar regime with limited observability and strong nonlinearity. The conditional flow architecture itself is not orbit-family-specific, but all reported training and testing data are drawn from NRHO trajectories. We have revised the abstract and added a dedicated limitations paragraph stating that generalization beyond NRHOs has not been demonstrated and would require retraining and validation on additional orbit classes. The core methodological contribution on coupling conditional density estimation with NLS refinement therefore remains scoped to the demonstrated regime.","revision_made":"yes","referee_comment":"[Abstract] Abstract / training description: the method is trained exclusively on perturbed topocentric observations from NRHO trajectories, yet the central claim of applicability to 'general angles-only cislunar OD' is not supported by any reported hold-out tests on other orbit families, different dynamical regimes, or realistic sensor noise models; if the conditional density overfits to NRHO-specific geometry and observability, the generated hypotheses will not remain accurate or reliable for NLS initialization outside the training distribution."}],"tokens_in":1253,"tokens_out":456,"duration_ms":24671,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The core of this paper is training a conditional normalizing flow on perturbed topocentric angles from Near Rectilinear Halo Orbits, then sampling the learned density to produce initial state guesses that get refined by nonlinear least squares. That pipeline is the main thing to know.\n\nWhat is actually new is the direct application of conditional flows to angles-only orbit determination in the cislunar regime, where short arcs can produce multimodal posteriors. The setup treats the problem as conditional density estimation and keeps the physics in the loop through the subsequent NLS step. That combination is a reasonable way to generate flexible, statistically consistent hypotheses without assuming a single Gaussian mode.\n\nThe paper does a clean job laying out the training on simulated NRHO data and the sampling-plus-refinement workflow. If the full manuscript includes the expected implementation details and loss formulation, that part is solid engineering.\n\nThe soft spot is the missing evidence on transfer. Everything is trained inside the NRHO family; there is no reported test on other cislunar orbit classes, different observation geometries, or real noise models. The central claim that the sampled states remain accurate and useful warm starts therefore rests on an untested assumption that the conditional density generalizes. The abstract also gives no error metrics, baseline comparisons, or hold-out results, so it is impossible to judge whether the method actually improves on existing approaches.\n\nThis is for people already working on generative methods for space situational awareness or lunar navigation who want to see normalizing flows tried on this specific problem. A reader looking for a ready-to-use algorithm will not get much yet.\n\nI would send it to peer review if the full paper adds quantitative validation on out-of-distribution cases; without that it is still too early.","headline":"Conditional normalizing flows on NRHO-simulated angles for cislunar OD, but no shown generalization beyond that narrow training set.","tokens_in":2211,"tokens_out":418,"would_cite":false,"duration_ms":7416,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"A conditional normalizing flow trained on perturbed NRHO observations learns flexible posteriors over initial states from short-arc angles-only measurements.","keywords":["normalizing flows","orbit determination","cislunar","angles-only","conditional density estimation","near rectilinear halo orbits","generative modeling"],"falsifier":"Test the trained flow on real angles-only observations from a cislunar spacecraft on a non-NRHO trajectory and measure whether the refined state estimates recover the true orbit within the uncertainty reported by the posterior.","tokens_in":2523,"feed_emoji":"🛰️","tokens_out":619,"duration_ms":20114,"temperature":0.7,"pith_summary":"The paper frames cislunar orbit determination as a conditional density estimation task solved with normalizing flows. The model is trained on sequences of topocentric angle observations generated from Near Rectilinear Halo Orbits under realistic perturbations. Once trained, the flow produces samples from the posterior distribution over possible initial spacecraft states given new measurements. These samples serve as statistically consistent starting points that are then polished by classical nonlinear least-squares. The result is a method that can represent uncertainty and multimodality while remaining anchored in the underlying orbital dynamics.","feed_headline":"Normalizing flow learns posterior for cislunar angles-only orbit determination","feed_subtitle":"Trained on perturbed NRHO data, the model samples consistent state hypotheses that warm-start least-squares refinement.","key_machinery":"Conditional normalizing flow performing density estimation of initial orbital states given short sequences of angles-only topocentric observations.","core_discovery":"A conditional normalizing flow trained on perturbed topocentric observations from Near Rectilinear Halo Orbits learns a flexible and potentially multimodal posterior over initial states conditioned on angles-only measurements; sampling from this density yields physics-informed state hypotheses that are refined by nonlinear least-squares to give competitive warm starts for classical orbit determination algorithms.","pith_inferences":["The same training strategy could be repeated for other cislunar orbit families or different sensor combinations to broaden coverage.","If the posterior quality holds under distribution shift, the approach may reduce reliance on long observation arcs or additional range data.","Embedding the flow in an onboard filter could support autonomous navigation with limited ground contact."],"forward_implications":["The learned density can represent multimodal posteriors when short-arc angles-only data leave multiple orbits consistent with the measurements.","State hypotheses sampled from the flow are statistically consistent with the training dynamics and measurement noise model.","The sampled hypotheses supply warm starts that improve convergence of nonlinear least-squares for angles-only problems.","Generative modelling is extended from simulation to a practical cislunar orbit determination pipeline."],"fun_headline_variants":["Conditional flows learn cislunar posteriors from angles-only data","Normalizing flows sample states for cislunar orbit refinement","Flows provide physics-informed warm starts in cislunar OD","Multimodal posteriors via flows for angles-only cislunar tasks"],"cache_read_input_tokens":64,"weakest_assumption_plain":"Training exclusively on perturbed topocentric observations from Near Rectilinear Halo Orbits produces a posterior that remains accurate and physics-consistent for real or out-of-distribution angles-only measurements in the cislunar regime.","fun_headline_variants_meta":{"raw":{"variants":["Conditional flows learn cislunar posteriors from angles-only data","Normalizing flows sample states for cislunar orbit refinement","Flows provide physics-informed warm starts in cislunar OD","Multimodal posteriors via flows for angles-only cislunar tasks"]},"model":"grok-4.3","cost_usd":0.003802,"raw_usage":{"total_tokens":1894,"prompt_tokens":531,"num_sources_used":0,"completion_tokens":69,"cost_in_usd_ticks":38024500,"prompt_tokens_details":{"text_tokens":531,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1294,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":531,"tokens_out":69,"duration_ms":10399,"temperature":1.0,"reasoning_tokens":1294,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-01T06:08:25.409838+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Test the trained flow on real angles-only observations from a cislunar spacecraft on a non-NRHO trajectory and measure whether the refined state estimates recover the true orbit within the uncertainty reported by the posterior.","supporting_citations":[],"review_version":1}