Open and completed research requests

This paper studies the theoretical foundations and system architecture of Quarks, with emphasis on multi-agent orchestration for long-horizon research tasks. The core idea is to model research execution as a contract-bounded computation graph where agents coordinate through typed interfaces, validators, and explicit phase transitions. We formalize roles for orchestrator, executor, and AI-scientist agents, including meta-orchestration policies for rerouting, backtracking, and self-correction under uncertainty. The framework combines persistent memory, prompt/template abstractions, and neural-symbolic control primitives to support reliable multi-stage reasoning over extended runtimes. Expected impact is a principled blueprint for robust “research-as-a-service” systems that are safe, robust, extensible, and empirically measurable.

Current neurosymbolic ASP pipelines require manual encoding redesign (for example carry-based formulations) to unlock treewidth-aware compilation gains. This project extends the dual-backend framework by automatically rewriting ASP constraints into semantically equivalent, lower-treewidth forms and dynamically routing each training instance to enumeration or compilation. The core idea is to combine static graph analysis of grounded programs with a semantics-preserving rewrite library and a learned crossover predictor based on treewidth, model-count proxies, and circuit reuse ratio. The expected impact is to remove expert-dependent encoding effort, broaden exact inference to more tasks, and improve training efficiency without sacrificing correctness.

We propose a reinforcement learning methodology that combines intrinsic novelty-driven exploration with explicit goal conditioning and extrinsic reward optimization. The core motivation is to avoid myopic exploitation while preserving path optimality for objective goals. The approach uses a dual-value design: one stream for task-optimal return and one for curiosity, with adaptive weighting that decays as goal confidence improves. We expect this to improve sample efficiency, robustness, and discovery of better trajectories in sparse- and deceptive-reward settings. The impact is a practical framework for agents that remain purpose-driven while retaining productive curiosity.

Continual learning systems must remain plastic enough to learn new tasks while stable enough to avoid catastrophic forgetting, but this tradeoff is increasingly constrained by memory behavior rather than compute alone. This project adopts a unified data-access perspective in which deterministic retrieval and stochastic sampling are treated within one framework, so memory support for both adaptation and retention can be analyzed jointly. We will design and evaluate memory mechanisms that explicitly manage stochastic demand, entropy pressure, and replay/update efficiency across task sequences. Expected impact is a clearer systems-level blueprint for continual learning stacks that improve retention and adaptation while remaining robust and efficient on realistic hardware.

This project targets a key safeguards challenge: identifying covert fissile production pathways in nuclear fusion reactor settings using antineutrino observations. We will study whether novel material configurations (blankets, moderators, shielding, and structural composites) create measurable antineutrino-spectrum or time-profile signatures that improve detectability. The approach combines physics-informed simulation of reactor scenarios with statistical detection models and material sensitivity analysis. We expect to produce a ranked set of material candidates and monitoring strategies that improve early-warning capability while quantifying uncertainty and false-alarm risk.

This project investigates whether real-world lottery draws deviate from ideal uniform randomness due to physical implementation effects such as ball weight, wear, machine mechanics, or procedural conditions. The approach combines number-theoretic feature design with statistical hypothesis testing and predictive modeling on historical draw data to detect persistent, reproducible structure. Rather than assuming dependence across draws, we explicitly test independence, stationarity, and mechanism-specific bias using both global and time-local analyses. The expected impact is a rigorous framework for distinguishing true mechanical bias from random fluctuation, with reproducible methods that can be applied across lottery systems.

Fourier transformation remains a core primitive in scientific computing, signal processing, and ML systems, but practical performance and accuracy depend heavily on algorithm and hardware choices. This project will build a reproducible benchmark framework to compare modern Fourier transformation implementations across representative workloads (1D/2D/3D, real/complex, varying sizes, and precision modes). We will evaluate both numerical fidelity and system-level efficiency to identify configuration-specific SOTA candidates rather than a one-size-fits-all winner. The expected impact is a decision guide and open benchmark suite that helps researchers and engineers select the best Fourier pipeline for their constraints.

We introduce SymbolicAI, a versatile and modular framework employing a logic-based approach to concept learning and flow management in generative processes. SymbolicAI enables the seamless integration of generative models with a diverse range of solvers by treating large language models (LLMs) as semantic parsers that execute tasks based on both natural and formal language instructions, thus bridging the gap between symbolic reasoning and generative AI. We leverage probabilistic programming principles to tackle complex tasks, and utilize differentiable and classical programming paradigms with their respective strengths. The framework introduces a set of polymorphic, compositional, and self-referential operations for multi-modal data that connects multi-step generative processes and aligns their outputs with user objectives in complex workflows. As a result, we can transition between the capabilities of various foundation models with in-context learning capabilities and specialized, fine-tuned models or solvers proficient in addressing specific problems. Through these operations based on in-context learning our framework enables the creation and evaluation of explainable computational graphs. Finally, we introduce a quality measure and its empirical score for evaluating these computational graphs, and propose a benchmark that compares various state-of-the-art LLMs across a set of complex workflows. We refer to the empirical score as the "Vector Embedding for Relational Trajectory Evaluation through Cross-similarity", or VERTEX score for short. The framework codebase and benchmark are linked below.