{"id":"0a81c64d-54b5-4eb2-ad26-0b54e792dc98","arxiv_id":"2611.17005","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A self-informed sampling rule that queries a time-optimal MPC expert only at policy-deviating states learns a neural controller with under half the data, matching the expert's region of attraction and enabling high-frequency control on a Furuta pendulum.","lead":"This paper trains a neural network to imitate a time-optimal model-predictive controller for a Furuta pendulum, using a new sampling rule that asks the expensive expert controller for training data only where the learned policy deviates from time-optimal behavior. This needs less than half the expert data of a standard baseline, matches the expert's success region, and runs hundreds of times faster, swinging up a real pendulum at 250 hertz.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"The data-efficiency claim compares baselines only until they match SIFS's failure rate (Fig. 5b caption), not to their own convergence; the abstract's 'matches standard dataset aggregation performance' is therefore not established.","rationale":"The paper is a coherent engineering integration with a real hardware swing-up (b182, b195) and a clear simulation comparison (b149). The reader's CONDITIONAL verdict is reasonable. My single most load-bearing concern is not the learned T2G per se, but the empirical comparison that carries the headline SIFS claim. The paper's own Fig. 5b caption says the baselines are measured by the queries/time 'needed to match the SIFS failure rate.' That makes the comparison asymmetric: SIFS is the target, and DAgger/MC-dropout are stopped at the target. The abstract and conclusion, however, state that SIFS 'matches standard dataset aggregation performance' (b4; conclusion). Matching a target set by the proposed method is not the same as matching the baseline's best performance. If DAgger's failure rate continues to fall after the match point, the correct statement is that SIFS is faster to a given intermediate failure rate, not that it matches DAgger with less data. The full SIFS training run in Section IV.C is 1441 iterations and 1,189,320 aggregated points; Fig. 5b's x-axis goes only to 240 iterations, and the paper does not report the baselines at the full budget. This is a concrete, checkable gap. The reader's T2G-oracle concern (b100) is a legitimate mechanism-level worry, and the unspecified control-invariant set C in Section III.F is a real missing premise for the formal constraint-satisfaction sentence. Both are worth fixing. But neither is as directly dispositive for the headline data-efficiency number as the truncated baseline comparison: a faithful T2G would still leave the abstract's 'matches' unsupported if the baselines were not run to convergence, while a proper baseline run would settle the data-efficiency claim regardless of the T2G explanation. The 1 kHz abstract versus 250 Hz hardware discrepancy (b5 versus b195) is a scope and communication issue, not the load-bearing one. The appropriate verdict remains CONDITIONAL, with the condition that the baseline comparison be run to the same budget or to convergence; if the baselines then match or beat SIFS, the central claim needs to be rewritten. I therefore disagree with the reader's specific assessment that the data-efficiency claim is already plausibly supported, and I would keep the verdict at CONDITIONAL but for this different reason.","tokens_in":15150,"tokens_out":17163,"duration_ms":102596,"concrete_test":"Run DAgger and MC-dropout for the same total expert-query budget as the full SIFS run in Section IV.C (1,189,320 aggregated points, 1441 iterations), or until their failure rates plateau. Compare final failure rates, final success sets, and the query/time at which each method first reaches SIFS's final failure rate. If either baseline reaches a final failure rate at or below SIFS's, the abstract's 'matches standard dataset aggregation performance with less than half the training data' is false and should be rewritten as 'reaches a given intermediate failure rate faster.' If both plateau above SIFS at the same budget, the data-efficiency claim survives.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central data-efficiency claim (abstract; b4; b169; conclusion) is that SIFS matches DAgger/MC-dropout with less than half the data and about one-third the training time. The only direct comparison is Fig. 5. Its panel (b) caption states that the reported queries and time are those 'needed to match the SIFS failure rate.' That is a one-sided stopping rule: the baselines are run only until they reach the failure level SIFS already has, not until they converge or until the same total query budget as the full SIFS run described in Section IV.C (1441 iterations, 1,189,320 aggregated points). If DAgger or MC-dropout keep improving beyond that match point, SIFS does not 'match' their performance; it merely reaches a particular intermediate failure rate faster. The paper does not report the baselines' final failure rates or identify the SIFS checkpoint used for the match, so the comparison cannot distinguish 'SIFS is more data-efficient' from 'SIFS converges to a worse final policy more quickly.' This is load-bearing because the novelty of SIFS rests on this comparison and because the baselines are the standard DAgger/MC-dropout methods, not weak controls. The text at b167-b169 reports a faster reduction and a match point, but never a baseline plateau. The reader's T2G-faithfulness concern is relevant to the mechanism, but even a faithful T2G would not rescue the claim if the baselines are truncated; the stopping rule must be fixed first.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper proposes an integrated pipeline for imitating a nonlinear time-optimal MPC (TOMPC) with hard terminal constraints by a neural network. The architecture combines an optimally sampled dataset, a time-to-go (T2G) based weighting of conflicting expert labels, a self-informed sampling (SIFS) strategy that queries the expert only at states where the learned T2G head deviates from the expected decrease of one sampling interval per step, barrier and penalty losses plus a safety filter for constraint satisfaction, and an ancillary DIC or LQR controller for model mismatch. The central claims are three: SIFS matches the failure rate of DAgger and MC-dropout with less than half the training data and about one-third of the training time; the final NNMPC matches the expert TOMPC's region of attraction (both controllers succeed in 91.5% of 1000 sampled initial states, with total success 95.7% vs 95.8%) while raising the reliable control frequency from about 1.36 Hz to up to 1 kHz on GPU; and the controller swings up and stabilizes a physical Furuta pendulum at 250 Hz along trajectories close to the expert solution.","tokens_in":15644,"tokens_out":28753,"duration_ms":176599,"significance":"The target problem is genuinely harder than the quadratic-tracking MPC imitation of most prior work, and a complete architecture with real-hardware validation would be a useful contribution if the claims hold. I credit the paper's strengths explicitly: a hardware demonstration on a Furuta pendulum with a long-horizon, nonconvex, hard-terminal-constraint TOMPC expert is rare; Table I gives a concrete and favorable computation-time comparison; the component ablations in Figs. 4 and 6 are good practice; the authors acknowledge expert suboptimality, learned-T2G noise, and remaining performance gaps; and the project webpage with videos and implementation details aids reproducibility. The obstacles are the strength of the evidence behind the two headline claims (data efficiency and region-of-attraction parity), the unspecified safe set behind the formal constraint-satisfaction guarantee, and the heuristic justification of the T2G selection signal. These are fixable within the manuscript's scope, so the appropriate outcome is a major revision rather than a rejection.","major_comments":[{"comment":"The headline data-efficiency claim rests on a one-sided comparison. The caption of Fig. 5b states that the reported queries and training times are those 'needed to match the SIFS failure rate', so DAgger and MC-dropout are stopped at the first point at which they reach a failure rate that SIFS has already achieved; the paper does not report where those baselines end up if trained to the full 1441-iteration budget of Section IV.C, and it does not identify the SIFS checkpoint (iteration, failure-rate value, query count) used as the match target. Under this stopping rule, the abstract's claim that SIFS 'matches standard dataset aggregation performance' cannot be distinguished from the alternative that SIFS reaches an intermediate failure rate faster while converging to a worse final policy. The comparison also needs internal reconciliation: the full SIFS run aggregates 1,189,320 expert-labeled points over 1441 iterations, while Fig. 5b's query axis tops out at 9,000 additional queries, so the match checkpoint and its failure-rate value must be stated explicitly. Please run the baselines to the full budget, report their final failure rates and the failure rate at the matching SIFS checkpoint, and state the match criterion before the comparison is made.","section":"Sec. VI.B (Fig. 5b); abstract; Sec. VIII","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Figs. 4 and 5 report a single failure-rate curve per method with no variance across training seeds, and the region-of-attraction comparison in Fig. 3 (both controllers succeed in 91.5% of the 1000 sampled initial states) is a single-network point estimate; the hardware evidence in Section VII is five runs of a single swing-up task. For claims whose magnitude is expressed as factors (about one-third of the training time, less than half the data, and 91.5% parity), seed-to-seed variance is essential. Please report means and standard deviations (or min-max envelopes) over at least three to five seeds for the failure-rate curves, query counts, and times-to-match, and state whether the hardware reruns vary the initial condition or only the execution noise.","section":"Figs. 3-5; Sec. VII","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The SIFS mechanism rests on the faithfulness of the learned T2G, and its justification at b100 is an informal plausibility claim. The argument that an exactly one-interval-per-step decrease is 'unlikely to occur by chance' is not quantified, and the paper's own acknowledgments of prediction noise and discontinuities (b88) and of expert suboptimality (b151) give concrete failure modes for the linear-trend template; the T2G labels are also capped by the N=60 horizon (1.8 s at dt=0.03), with no stated convention for the case where the expert's predicted trajectory does not enter the xi-ball within the horizon. Because the T2G is learned from expert labels and then used both to select query states and to weight conflicting labels, the loop is a bootstrap rather than a formal reduction; the central question is whether flags actually coincide with policy suboptimality. A concrete diagnostic would settle this: report the correlation between flagged states and one-step imitation error or closed-loop failure, or compare SIFS with a selection rule based on the ground-truth expert T2G residual, while holding the query budget fixed.","section":"Sec. III.D; Sec. IV.A","authors_acknowledge":"Since the T2G is learned, its predictions contain noise and occasional harmless discontinuities.","authors_acknowledge_p":0.69},{"comment":"The formal constraint-satisfaction guarantee of Section III.F is conditional on a control-invariant set C that is never instantiated for the Furuta plant. The section acknowledges that the maximal control-invariant set may be intractable and offers CBFs as an alternative, but the paper does not state which option the experiments use, whether the projection onto U_s(x) is actually computed online at 250 Hz, or whether the base-angle and input box constraints were violated in the 1000-state evaluation of Fig. 3 or in the hardware runs. Consequently, the claim that the safety filter 'formally guarantees constraint satisfaction' is not verifiable for the demonstrated system. Please specify the concrete C (or the CBF and its relative-degree assumptions) used in the Furuta example, state how the projection's differentiability is handled in the training loss (9), and report constraint-violation statistics for the simulation and hardware evaluations.","section":"Sec. III.F; Secs. IV-V","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"As printed, the exponent in Eq. (2) contains T2G(x_i) - T2G(x_i), which is identically zero, so the weight is constant and the equation cannot be the mechanism behind the ablation in Fig. 4; please correct the formula (presumably T2G(x_i) minus the neighbor-averaged T2G of Eq. (1)) and state the sign convention so that the most time-optimal samples receive the largest weight.","section":"Sec. III.C, Eq. (2)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Please clarify what 'training time' includes (network retraining, expert queries, or both), whether the query counts are additional to the 15,601-point initial grid of Section IV.C, and what failure-rate value is attained at the match point.","section":"Fig. 5b","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The symbol k denotes both the MPC time step in Eq. (8) and the number of neighbors in the weight notation w_knn of Eq. (9); please rename one of the two to avoid confusion.","section":"Sec. IV.B","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The perturbation table is difficult to parse; please label the rows with the perturbation value, the columns with the controller variant, and use a consistent color scale so that the success rates are legible.","section":"Fig. 6","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Please state how the representative hardware run was chosen and whether the four additional reruns share the same initial condition, since this determines how much generality the hardware claim carries.","section":"Sec. VII and Fig. 7","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The abstract's '1 kHz' is the GPU-inference bound from Table I, whereas the deployed hardware controller runs at 250 Hz; please make that distinction explicit.","section":"Sec. V.B and Sec. VIII","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Please state whether the deployed pipeline includes the ancillary controller and the safety filter of Fig. 1, and whether the safety filter is active during the hardware runs.","section":"Sec. VII","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Section IV.C states that the iterative procedure runs for 1441 iterations while Fig. 5a shows only 900; please state whether the figure truncates the run and whether the region-of-attraction evaluation in Fig. 3 uses the final iteration-1441 network.","section":"Sec. IV.C and Fig. 5a","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision"},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"The paper's core idea—self-informed sampling (SIFS) using a jointly learned time-to-go as an informative-state oracle—is genuinely original. The T2G-based weighting for conflicting expert labels is also new and lightweight. The integrated architecture is a sensible assembly of existing building blocks, and the hardware validation on a Furuta pendulum, while limited, is a real step beyond the sim-only work that dominates this area. The authors are honest about limitations: they acknowledge T2G noise, suboptimal expert labels, and the remaining performance gap. The writing is clear and the method is coherently explained.\n\nThe soft spot is load-bearing. The stress-test note is correct: the data-efficiency comparison in Fig. 5b uses a one-sided stopping rule. DAgger and MC-dropout are run only until they match SIFS's failure rate, not until they converge or exhaust the same query budget. The paper never reports the baselines' final failure rates or plateaus. This means the claim that SIFS 'matches standard dataset aggregation performance with less than half the data and a third of the time' is not established by the evidence shown. It could be that SIFS converges to a worse final policy faster. The authors need to show the full learning curves for all methods, not just the match point.\n\nOther concerns are real but less severe. The hardware demonstration is five runs of a single swing-up task—encouraging but not statistically robust. The safety filter's control-invariant set is never specified for the Furuta plant, so the formal constraint-satisfaction guarantee is unverified. Several hyperparameters are unreported. The abstract claims 1 kHz, but hardware runs at 250 Hz. These are addressable in revision.\n\nWho this is for: researchers working on neural approximation of MPC, especially for time-optimal or computationally expensive formulations. The paper deserves a serious referee—the core idea is worth engaging with—but the referee should push hard on the stopping-rule issue and ask for full learning curves. I would not cite it in its current form, but I would read a revised version carefully.","headline":"SIFS is a genuinely new sampling idea for NNMPC, but the central data-efficiency claim is undercut by a one-sided stopping rule in the baseline comparison.","tokens_in":16231,"tokens_out":525,"duration_ms":9695},"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek/deepseek-v4-flash","headline":"Neural network imitates time-optimal MPC at 1 kHz on a real pendulum","keywords":["neural network model predictive control","time-optimal control","imitation learning","self-informed sampling","time-to-go","Furuta pendulum","data-efficient expert sampling","constraint satisfaction"],"falsifier":"Run SIFS on a system where the expert MPC is known to be suboptimal (e.g., stuck in a local minimum) and the T2G of the expert is itself inconsistent; if SIFS still identifies informative states based on deviations from the expected T2G trend, it may incorrectly flag states where the NNMPC is actually better than the expert, undermining the assumption that T2G deviation indicates poor control.","tokens_in":16601,"feed_emoji":"🤖","tokens_out":1784,"duration_ms":15807,"temperature":0.7,"pith_summary":"This paper presents an integrated architecture for training a neural network to imitate a computationally expensive time-optimal model predictive controller (TOMPC) and deploying it in real time on a physical Furuta pendulum. The key claim is that the proposed self-informed sampling (SIFS) strategy, which uses the network's own learned time-to-go prediction to identify states where the expert controller should be queried, matches the performance of standard data aggregation (DAgger) with less than half the training data and about one-third the training time. The resulting neural controller achieves a 91.5% success rate over 1000 random initial states, matching the expert's region of attraction, while raising the control frequency from 1.4 Hz to up to 1 kHz. The paper further demonstrates real-world swing-up and stabilization on hardware, with trajectories closely matching simulation. If correct, this work shows that neural imitation can make long-horizon, nonconvex TOMPC with hard terminal constraints deployable on fast dynamical systems, overcoming the data-generation bottleneck that previously limited such approaches.","feed_headline":"Neural net mimics time-optimal MPC at 1 kHz on a real pendulum","feed_subtitle":"Self-informed sampling cuts expert data by half; hardware swing-up matches simulation.","key_machinery":"The central mechanism is the jointly learned time-to-go (T2G) output head of the neural network. The T2G is defined as the first time the predicted MPC trajectory from a state lies within a tolerance of the target state. Along a time-optimal trajectory, the T2G decreases by exactly one sampling interval per step. The self-informed sampling (SIFS) algorithm monitors the T2G trend over entire rollouts: if the T2G deviates persistently from this expected linear decrease (e.g., stagnation, increases, or irregular jumps), the states at that transition are flagged as informative and the expert MPC is queried there. This serves as a direct, dynamics-aware indicator of control quality, unlike indirect proxies such as epistemic uncertainty or trajectory distance. The paper also employs a lightweight T2G-based weighting scheme for conflicting expert labels (Section III-C), optimal sampling to reduce dataset correlation (Section III-B), a safety filter with logarithmic barrier and quadratic penalty in the loss (Section III-F), and ancillary controllers (DIC and LQR) for plant-model mismatch compensation (Section III-E).","core_discovery":"The paper claims that a neural network can be trained to imitate a long-horizon, nonconvex time-optimal model predictive controller (TOMPC) with hard terminal constraints, and that the resulting neural network model predictive controller (NNMPC) can be deployed in real time on a fast dynamical system (a Furuta pendulum) while matching the expert controller's region of attraction. The central discovery is that self-informed sampling (SIFS), which uses a jointly learned time-to-go (T2G) value function to identify states where the policy deviates from time-optimal behavior, drastically reduces the amount of expert data required for training. The authors argue that a T2G prediction that decreases by exactly one sampling interval per time step is unlikely to occur by chance unless the network has learned both the T2G of visited states and the controls that steer the system to the correct next state, making it a direct and reliable indicator of control quality. The paper reports that SIFS matches standard dataset aggregation (DAgger) performance in roughly one-third the training time and with less than half the training data, and that the resulting NNMPC achieves a 91.5% success rate over 1000 sampled initial states (identical to the expert MPC), while raising the control frequency from 1.4 Hz to up to 1 kHz. On hardware, the controller swings up and stabilizes the pendulum along trajectories closely matching simulation.","pith_inferences":["The SIFS criterion's reliance on a learned T2G that decreases by exactly one sampling interval per step could be adapted to other iterative optimal control problems where the cost-to-go satisfies a Bellman-like recursion, not just time-optimal formulations.","The paper's hardware validation on a Furuta pendulum, while convincing, is a single test case; the scalability of SIFS to higher-dimensional systems (e.g., quadrotors, manipulators) where the T2G may be harder to learn or where the state space is larger remains to be demonstrated.","The authors acknowledge that conflicting labels from equally time-optimal modes remain unresolved by their weighting scheme; a future extension could use mixture density networks or multi-modal architectures to explicitly handle such cases.","The safety filter relies on a control-invariant safe set C that is not explicitly specified for the Furuta plant; the practical constraint-satisfaction guarantee therefore depends on the conservativeness of the chosen CBF or invariant set, which may limit performance in more complex environments."],"forward_implications":["Neural imitation of time-optimal MPC with hard terminal constraints becomes feasible for fast dynamical systems where the expert's computational cost previously limited both online deployment and training-data generation.","The SIFS criterion can be applied to any control problem with a natural value function that decreases linearly along optimal trajectories, potentially extending beyond time-optimal objectives to other monotonically decreasing cost-to-go functions.","The integrated architecture demonstrates that dataset correlation, suboptimal labels, covariate shift, constraint satisfaction, and model mismatch can be jointly addressed in a single training and deployment pipeline, reducing the fragmentation of existing NNMPC approaches.","The 700-fold reduction in worst-case computation time (from 732 ms to 1 ms) opens the door to deploying neural approximations of complex MPC formulations on embedded hardware with limited computational resources."],"supporting_citations":[{"why":"Provides the standard DAgger algorithm that SIFS is compared against and aims to improve upon in data efficiency.","marker":"[11]"},{"why":"Supplies the optimally-sampled dataset algorithm used to reduce dataset correlation in the proposed architecture.","marker":"[12]"},{"why":"Provides the MC-dropout based uncertainty-aware data aggregation method that serves as a baseline comparison for SIFS.","marker":"[19]"},{"why":"Describes the exponentially weighted objective used to realize time-optimality in the expert MPC formulation.","marker":"[2]"},{"why":"Fatrop solver is used to transcribe and solve the optimal control problem for the expert MPC.","marker":"[31]"},{"why":"Provides the continuous-time dynamics model of the Furuta pendulum used in the MPC and simulation.","marker":"[30]"},{"why":"Control barrier functions are cited as a conservative alternative for the safety filter when the maximal control-invariant set is intractable.","marker":"[25]"}],"fun_headline_variants":[],"cache_read_input_tokens":0,"weakest_assumption_plain":"The self-informed sampling (SIFS) strategy assumes that a neural network that cannot decrement an internal counter will only produce a time-to-go that decreases by exactly one sampling interval per step if it has genuinely learned the T2G of visited states and predicts controls that steer to the correct next state; if the learned T2G is a noisy or biased oracle, the criterion will flag the wrong states and the data-efficiency claim collapses.","fun_headline_variants_meta":{"raw":{"variants":[]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":0.000103,"raw_usage":{"cost":0.0001025752,"is_byok":false,"cost_details":{"upstream_inference_cost":0.0001025752,"upstream_inference_prompt_cost":8.67412e-05,"upstream_inference_completions_cost":1.5834e-05},"total_tokens":1245,"prompt_tokens":1158,"completion_tokens":87,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":256,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":0}},"tokens_in":1158,"tokens_out":87,"cost_source":"openrouter_usage","duration_ms":3874,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":0,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-13T08:02:20.562978+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"pipeline_version":"pith-pipeline@v1.4.1-daily-deepseek-or-medium","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"Run SIFS on a system where the expert MPC is known to be suboptimal (e.g., stuck in a local minimum) and the T2G of the expert is itself inconsistent; if SIFS still identifies informative states based on deviations from the expected T2G trend, it may incorrectly flag states where the NNMPC is actually better than the expert, undermining the assumption that T2G deviation indicates poor control.","supporting_citations":[{"cited_title":"A reduction of imitation learning and structured prediction to no-regret online learning,","cited_arxiv_id":null,"evidence_quote":"Provides the standard DAgger algorithm that SIFS is compared against and aims to improve upon in data efficiency."},{"cited_title":"Neural networks for on- chip model predictive control: A method to build optimized training datasets and its application to type-1 diabetes,","cited_arxiv_id":null,"evidence_quote":"Supplies the optimally-sampled dataset algorithm used to reduce dataset correlation in the proposed architecture."},{"cited_title":"Uncertainty-aware data aggregation for deep imitation learning,","cited_arxiv_id":null,"evidence_quote":"Provides the MC-dropout based uncertainty-aware data aggregation method that serves as a baseline comparison for SIFS."},{"cited_title":"Practical approaches for time- optimal mpc: Dual stage and artificial reference methods,","cited_arxiv_id":null,"evidence_quote":"Describes the exponentially weighted objective used to realize time-optimality in the expert MPC formulation."},{"cited_title":"Fatrop: A fast constrained optimal control problem solver for robot trajectory optimization and control,","cited_arxiv_id":null,"evidence_quote":"Fatrop solver is used to transcribe and solve the optimal control problem for the expert MPC."},{"cited_title":"On the dynamics of the furuta pendulum,","cited_arxiv_id":null,"evidence_quote":"Provides the continuous-time dynamics model of the Furuta pendulum used in the MPC and simulation."},{"cited_title":"Control barrier function based quadratic programs for safety critical systems,","cited_arxiv_id":null,"evidence_quote":"Control barrier functions are cited as a conservative alternative for the safety filter when the maximal control-invariant set is intractable."}],"lane_accuracy":{"status":"not yet measured","lane":{"model":"deepseek/deepseek-v4-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}