GVGoli Vaisi

Goli Vaisi

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Goli Vaisi has a total of 8 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • GPS–Relational: Generative Problem Solving with Relational Databases

    by Goli Vaisi, Phillip C-Y Sheu

    Large language models (LLMs) have improved natural language to SQL translation, but evaluating their outputs remains difficult when queries become complex, deeply nested, or produced at scale. Existing evaluations often rely on standard benchmarks with limited structural complexity or on model-based judgments that may be inconsistent. We present GPS-Relational, an evaluation framework that validates LLM-generated SQL by constructing an executable reference SQL query from the same natural-language intent and database schema. GPS-Relational builds this reference through grammar-guided decomposition, schema mapping, semantic validation, and SQL assembly. The reference SQL is executed to produce the expected result, which is then compared with the execution result of the LLM-generated SQL, allowing equivalent SQL forms to be accepted without requiring textual or structural matching. Additionally, GPS-Relational facilitates large-scale benchmark generation by systematically synthesizing extensive datasets of natural-language and SQL pairs with controlled nesting depth and relational complexity. These generated cases enable a systematic methodology to evaluate LLM performance across varied levels of structural complexity. Experimental results demonstrate that GPS-Relational constructs executable reference SQL across existing benchmark and complex generated queries, and that LLM execution accuracy degrades significantly as nesting depth increases up to 15 levels.

  • Open Access

    Virtual Mice, Real Errors: A Sensor-Aware Generative Framework for In Silico Ethology

    by Reza Sayfoori, Goli Vaisi, Hung Cao

    Long-duration animal trajectories are central to computational ethology, yet constructing large rodent cohorts remains costly, time-intensive, and constrained by animal-use considerations. We present a sensor-aware generative framework that separates latent behavioral dynamics from sensing-induced observation distortion to synthesize observed-domain trajectories that are behaviorally plausible while reproducing proxy-referenced observation distortions. The framework combines a run-level semi-Markov ethology model, occupancy calibration, and state-conditioned kinematic generation with a regime-dependent Ultra-Wideband observation channel that explicitly captures Line-of-Sight and Non-Line-of-Sight sensing conditions. Using four UWB sessions, this proof-of-concept study models three states—exploring, feeding, and burrowing—and evaluates realism through state occupancy, state-conditioned kinematic divergence, residual-domain agreement, and mean-squared displacement across time lags. We further assess whether sensor-aware conditioning improves robustness under LoS/NLoS domain shift in downstream trajectory classification. Sensor-aware conditioning yields stable mixed-domain performance with AUC = 0.995, whereas condition-agnostic baselines decline to AUC = 0.974 and AUC = 0.901. These results support the feasibility of sensor-aware in silico ethology as a proof-of-concept framework for controlled robustness studies and algorithm evaluation under proxy-referenced observation distortion. Because the present evaluation is based on four UWB sessions and uses a smoothed UWB-derived reference trajectory rather than independent ground truth, broader applications to synthetic-cohort generation, disease modeling, and statistical power-analysis workflows should be considered future directions requiring validation in larger datasets.

  • Open Access

    Extracting Forward Invariant Sets from Neural Network-Based Control Barrier Functions

    by Goli Vaisi, James Ferlez, Yasser Shoukry

    Training Neural Networks (NNs) to serve as Barrier Functions (BFs) is a popular way to improve the safety of autonomous dynamical systems. Despite significant practical success, these methods are not generally guaranteed to produce true BFs in a provable sense, which undermines their intended use as safety certificates. In this paper, we consider the problem of formally certifying a learned NN as a BF with respect to state avoidance for an autonomous system: viz. computing a region of the state space on which the candidate NN is provably a BF. In particular, we propose a sound algorithm that efficiently produces such a certificate set for a shallow NN. Our algorithm combines two novel approaches: it first uses NN reachability tools to identify a subset of states for which the output of the NN does not increase along system trajectories; then, it uses a novel enumeration algorithm for hyperplane arrangements to find the intersection of the NN's zero-sub-level set with the first set of states. In this way, our algorithm soundly finds a subset of states on which the NN is certified as a BF. We further demonstrate the effectiveness of our algorithm at certifying for real-world NNs as BFs in two case studies. We complemented these with scalability experiments that demonstrate the efficiency of our algorithm.

  • Open Access

    EnergyShield: Provably-Safe Offloading of Neural Network Controllers for Energy Efficiency

    by Mohanad Odema, James Ferlez, Goli Vaisi, Yasser Shoukry, Mohammad Abdullah Al Faruque

    To mitigate the high energy demand of Neural Network (NN) based Autonomous Driving Systems (ADSs), we consider the problem of offloading NN controllers from the ADS to nearby edge-computing infrastructure, but in such a way that formal vehicle safety properties are guaranteed. In particular, we propose the EnergyShield framework, which repurposes a controller "shield" as a low-power runtime safety monitor for the ADS vehicle. Specifically, the shield in EnergyShield provides not only safety interventions but also a formal, state-based quantification of the tolerable edge response time before vehicle safety is compromised. Using EnergyShield, an ADS can then save energy by wirelessly offloading NN computations to edge computers, while still maintaining a formal guarantee of safety until it receives a response (on-vehicle hardware provides a just-in-time fail safe). To validate the benefits of EnergyShield, we implemented and tested it in the Carla simulation environment. Our results show that EnergyShield maintains safe vehicle operation while providing significant energy savings compared to on-vehicle NN evaluation: from 24% to 54% less energy across a range of wireless conditions and edge delays.