arXiv subjects

Damian Dailisan

Publications and source records attributed to Damian Dailisan.

8 recordsLinked to original sources

arXiv2026

GROSS: German Rail Open-Source SUMO Scenario

Microscopic simulation enables reproducible evaluation in intelligent transportation systems, yet most open SUMO scenarios and toolchains remain road-traffic centric, leaving rail underrepresented despite its importance for public transport and its sensitivity to network-wide disruptions. We present the German Rail Open-Source Scenario (GROSS), an open pipeline that combines OpenStreetMap railway infrastructure with GTFS schedules to generate nation-scale rail scenarios for SUMO (Simulation of Urban MObility). Existing conversions often rely on geometry-only stop-to-track matching and inconsistent platform/track assignments, which can create routing anomalies and unstable simulations dominated by teleportation artefacts. GROSS addresses this with topology-aware stop mapping via a hierarchical station model, followed by station-level routing with validation and targeted repair. Across multiple German regions, GROSS reduces average teleportations per vehicle by a factor of 1.7--76.8$\times$, shortens delays compared to the vanilla SUMO pipeline, and it enables end-to-end generation of a Germany-wide scenario with 35\,925 trips for comparisons with operator-reported delay statistics. While the remaining long delays highlight limitations in available timetable metadata and rail dispatch modeling, GROSS lowers the barrier to building scalable, fully open rail simulations and to studying delay propagation at country scale.

cs.CE↗

arXiv2026

Belief Engine: Configurable and Inspectable Stance Dynamics in Multi-Agent LLM Deliberation

LLM-based agents are increasingly used to simulate deliberative interactions such as negotiation, conflict resolution, and multi-turn opinion exchange. Yet generated transcripts often do not reveal why an agent's stance changes: movement may reflect evidence uptake, anchoring, role drift, echoing, or changed prompt and retrieval context. We introduce the Belief Engine (BE), an auditable belief-update layer that treats "belief" as an evidential state over a proposition and exposes it as scalar stance. BE extracts arguments into structured memory and updates stance with a log-odds rule controlled by evidence uptake u and prior anchoring a. Across multiple base LLMs, parameter sweeps show that these controls reliably shape stance dynamics while preserving an evidence-level update trail. On DEBATE, a human deliberation dataset with pre/post opinions, BE best reconstructs participants whose final stance follows extracted evidence; stable and evidence-opposed cases instead point to anchoring or factors outside the extracted evidence stream. BE provides configurable infrastructure for studying evidence-grounded deliberation, where openness, commitment, convergence, and disagreement can be tied to explicit update assumptions rather than hidden prompt effects.

cs.AI↗

arXiv2025

Addressing Moral Uncertainty using Large Language Models for Ethical Decision-Making

We present an ethical decision-making framework that refines a pre-trained reinforcement learning (RL) model using a task-agnostic ethical layer. Following initial training, the RL model undergoes ethical fine-tuning, where human feedback is replaced by feedback generated from a large language model (LLM). The LLM embodies consequentialist, deontological, virtue, social justice, and care ethics as moral principles to assign belief values to recommended actions during ethical decision-making. An ethical layer aggregates belief scores from multiple LLM-derived moral perspectives using Belief Jensen-Shannon Divergence and Dempster-Shafer Theory into probability scores that also serve as the shaping reward, steering the agent toward choices that align with a balanced ethical framework. This integrated learning framework helps the RL agent navigate moral uncertainty in complex environments and enables it to make morally sound decisions across diverse tasks. Our approach, tested across different LLM variants and compared with other belief aggregation techniques, demonstrates improved consistency, adaptability, and reduced reliance on handcrafted ethical rewards. This method is especially effective in dynamic scenarios where ethical challenges arise unexpectedly, making it well-suited for real-world applications.

cs.CY↗

arXiv2025

Overcoming the Price of Anarchy by Steering with Recommendations

Varied real world systems such as transportation networks, supply chains and energy grids present coordination problems where many agents must learn to share resources. It is well known that the independent and selfish interactions of agents in these systems may lead to inefficiencies, often referred to as the `Price of Anarchy'. Effective interventions that reduce the Price of Anarchy while preserving individual autonomy are of great interest. In this paper we explore recommender systems as one such intervention mechanism. We start with the Braess Paradox, a congestion game model of a routing problem related to traffic on roads, packets on the internet, and electricity on power grids. Following recent literature, we model the interactions of agents as a repeated game between $Q$-learners, a common type of reinforcement learning agents. This work introduces the Learning Dynamic Manipulation Problem, where an external recommender system can strategically trigger behavior by picking the states observed by $Q$-learners during learning. Our computational contribution demonstrates that appropriately chosen recommendations can robustly steer the system towards convergence to the social optimum, even for many players. Our theoretical and empirical results highlight that increases in the recommendation space can increase the steering potential of a recommender system, which should be considered in the design of recommender systems.

cs.GT↗

arXiv2024

LLM Voting: Human Choices and AI Collective Decision Making

This paper investigates the voting behaviors of Large Language Models (LLMs), specifically GPT-4 and LLaMA-2, their biases, and how they align with human voting patterns. Our methodology involved using a dataset from a human voting experiment to establish a baseline for human preferences and conducting a corresponding experiment with LLM agents. We observed that the choice of voting methods and the presentation order influenced LLM voting outcomes. We found that varying the persona can reduce some of these biases and enhance alignment with human choices. While the Chain-of-Thought approach did not improve prediction accuracy, it has potential for AI explainability in the voting process. We also identified a trade-off between preference diversity and alignment accuracy in LLMs, influenced by different temperature settings. Our findings indicate that LLMs may lead to less diverse collective outcomes and biased assumptions when used in voting scenarios, emphasizing the need for cautious integration of LLMs into democratic processes.

cs.CL↗

arXiv2023

Dynamic value alignment through preference aggregation of multiple objectives

The development of ethical AI systems is currently geared toward setting objective functions that align with human objectives. However, finding such functions remains a research challenge, while in RL, setting rewards by hand is a fairly standard approach. We present a methodology for dynamic value alignment, where the values that are to be aligned with are dynamically changing, using a multiple-objective approach. We apply this approach to extend Deep $Q$-Learning to accommodate multiple objectives and evaluate this method on a simplified two-leg intersection controlled by a switching agent.Our approach dynamically accommodates the preferences of drivers on the system and achieves better overall performance across three metrics (speeds, stops, and waits) while integrating objectives that have competing or conflicting actions.

cs.AI↗

arXiv2022

Adaptive Coordination Offsets for Signalized Arterial Intersections using Deep Reinforcement Learning

Coordinating intersections in arterial networks is critical to the performance of urban transportation systems. Deep reinforcement learning (RL) has gained traction in traffic control research along with data-driven approaches for traffic control systems. To date, proposed deep RL-based traffic schemes control phase activation or duration. Yet, such approaches may bypass low volume links for several cycles in order to optimize the network-level traffic flow. Here, we propose a deep RL framework that dynamically adjusts offsets based on traffic states and preserves the planned phase timings and order derived from model-based methods. This framework allows us to improve arterial coordination while maintaining phase order and timing predictability. Using a validated and calibrated traffic model, we trained the policy of a deep RL agent that aims to reduce travel delays in the network. We evaluated the resulting policy by comparing its performance against the phase offsets deployed along a segment of Huntington Drive in the city of Arcadia. The resulting policy dynamically readjusts phase offsets in response to changes in traffic demand. Simulation results show that the proposed deep RL agent outperformed the baseline on average, effectively reducing delay time by 13.21% in the AM Scenario, 2.42% in the Noon scenario, and 6.2% in the PM scenario when offsets are adjusted in 15-minute intervals. Finally, we also show the robustness of our agent to extreme traffic conditions, such as demand surges in off-peak hours and localized traffic incidents

eess.SY↗

arXiv2022

Deep-learned orthogonal basis patterns for fast, noise-robust single-pixel imaging

Single-pixel imaging (SPI) is a novel, unconventional method that goes beyond the notion of traditional cameras but can be computationally expensive and slow for real-time applications. Deep learning has been proposed as an alternative approach for solving the SPI reconstruction problem, but a detailed analysis of its performance and generated basis patterns when used for SPI is limited. We present a modified deep convolutional autoencoder network (DCAN) for SPI on 64x64 pixel images with up to 6.25% compression ratio and apply binary and orthogonality regularizers during training. Training a DCAN with these regularizers allows it to learn multiple measurement bases that have combinations of binary or non-binary, and orthogonal or non-orthogonal patterns. We compare the reconstruction quality, orthogonality of the patterns, and robustness to noise of the resulting DCAN models to traditional SPI reconstruction algorithms (such as Total Variation minimization and Fourier Transform). Our DCAN models can be trained to be robust to noise while still having fast enough reconstruction times (~3 ms per frame) to be viable for real-time imaging.

eess.IV↗