arXiv subjects

Shashank Prakash

Publications and source records attributed to Shashank Prakash.

2 recordsLinked to original sources

arXiv2026

Nishpaksh: TEC Standard-Compliant Framework for Fairness Auditing and Certification of AI Models

The growing reliance on Artificial Intelligence (AI) models in high-stakes decision-making systems, particularly within emerging telecom and 6G applications, underscores the urgent need for transparent and standardized fairness assessment frameworks. While global toolkits such as IBM AI Fairness 360 and Microsoft Fairlearn have advanced bias detection, they often lack alignment with region-specific regulatory requirements and national priorities. To address this gap, we propose Nishpaksh, an indigenous fairness evaluation tool that operationalizes the Telecommunication Engineering Centre (TEC) Standard for the Evaluation and Rating of Artificial Intelligence Systems. Nishpaksh integrates survey-based risk quantification, contextual threshold determination, and quantitative fairness evaluation into a unified, web-based dashboard. The tool employs vectorized computation, reactive state management, and certification-ready reporting to enable reproducible, audit-grade assessments, thereby addressing a critical post-standardization implementation need. Experimental validation on the COMPAS dataset demonstrates Nishpaksh's effectiveness in identifying attribute-specific bias and generating standardized fairness scores compliant with the TEC framework. The system bridges the gap between research-oriented fairness methodologies and regulatory AI governance in India, marking a significant step toward responsible and auditable AI deployment within critical infrastructure like telecommunications.

cs.CY↗

arXiv2024

Tagged particle behavior in a harmonic chain of direction reversing active Brownian particles

We study the tagged particle dynamics in a harmonic chain of direction reversing active Brownian particles, with spring constant $k$, rotation diffusion coefficient $D_{\text{r}}$, and directional reversal rate $\gamma$. We exactly compute the tagged particle position variance for quenched and annealed initial orientations of the particles. For well-separated time scales, $k^{-1}$, $D_{\text{r}}^{-1}$ and $\gamma^{-1}$, the strength of spring constant $k$ relative to $D_{\text{r}}$ and $\gamma$ gives rise to different coupling limits and for each coupling limit there are short, intermediate, and long time regimes. In the thermodynamic limit, we show that, to the leading order, the tagged particle variance exhibits an algebraic growth $t^{\nu}$, where the value of the exponent $\nu$ depends on the specific regime. For a quenched initial orientation, the exponent $\nu$ crosses over from $3$ to $1/2$, via intermediate values $5/2$ or $1$, depending on the specific coupling limits. On the other hand, for the annealed initial orientation, $\nu$ crosses over from $2$ to $1/2$ via an intermediate value $3/2$ or $1$ for strong coupling limit and weak coupling limit respectively. An additional time scale $t_N=N^2/k$ emerges for a system with a finite number of oscillators $N$. We show that the behavior of the tagged particle variance across $t_N$ can be expressed in terms of a crossover scaling function, which we find exactly. Finally, we characterize the stationary state behavior of the separation between two consecutive particles by calculating the corresponding spatio-temporal correlation function.

cond-mat.stat-mech↗