Anannya Ghosh Tusti

Anannya Ghosh Tusti

Auto-generated profile, based on publicly available information

Anannya Ghosh Tusti

Texas State University

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
4 February 2026 in ACM Computing Surveys
Verified
Journal Article
Open access
Shriyank SomvanshiShriyank Somvanshi
Shriyank Somvanshi
MIMonzurul Islam
Monzurul Islam
ARAmir Rafe
Amir Rafe
ATAnannya Ghosh Tusti
Anannya Ghosh Tusti
ACArka Chakraborty
Arka Chakraborty
SDSubasish Das
Subasish Das
Mechanistic interpretability seeks to reverse-engineer the internal logic of neural networks by uncovering human-understandable circuits, algorithms, and causal structures that drive model behavior. Unlike post hoc explanations that describe what models do, this paradigm focuses on why and how they compute, tracing information flow through neurons, attention heads, and activation pathways. This survey provides a high-level synthesis of the field-highlighting its motivation, conceptual foundations, and methodological taxonomy rather than enumerating individual techniques. We organize mechanistic interpretability across three abstraction layers— neurons , circuits , and algorithms —and three evaluation perspectives: behavioral , counterfactual , and causal . We further discuss representative approaches and toolchains that enable structural analysis of modern AI systems, outlining how mechanistic interpretability bridges theoretical insights with practical transparency. Despite rapid progress, challenges persist in scaling these analyses to frontier models, resolving polysemantic representations, and establishing standardized causal benchmarks. By connecting historical evolution, current methodologies, and emerging research directions, this survey aims to provide an integrative framework for understanding how mechanistic interpretability can support transparency, reliability, and governance in large-scale AI.


6 May 2025 in Traffic Injury Prevention
Verified
Journal Article
Closed access
RCRohit Chakraborty
Rohit Chakraborty
ATAnannya Ghosh Tusti
Anannya Ghosh Tusti
AHAhmed Hossain
Ahmed Hossain
MSMohaddese Salehian
Mohaddese Salehian
SJSyed Aaqib Javed
Syed Aaqib Javed
SDSubasish Das
Subasish Das
This study aims to identify key contributing factors to driver ejection in vehicle crashes. By analyzing crash characteristics, restraint usage, and environmental conditions, the study provides data-driven insights to mitigate ejection risks and improve roadway safety interventions. This study, utilizing comprehensive Texas crash data from 2017 to 2022; investigates this critical concern by applying an ARM algorithm to uncover key patterns and interactions among various factors contributing to ejection events. By analyzing the specific circumstances under which driver ejections occur, both with and without restraint usage, the study examines the dynamics of these crashes and their contributing factors. The analysis reveals that most fatal and severe ejection crashes involving restrained drivers are strongly associated with vehicle overturns, compounded by the interaction of dark lighting conditions, curved road segments, high posted speed limits (70 mph or higher), and two-lane, two-way rural roads. In contrast, ejection crashes involving unrestrained drivers are predominantly linked to dark lighting conditions, rural two-lane roads, and pickup or SUV-type vehicles, where the absence of restraint significantly increases the risk of fatal or severe injuries. Airbag deployment status, particularly cases of non-deployment or multiple deployments, has been observed in combination with other crash factors, such as high-speed impacts and overturning, contributing to an increased likelihood of ejection. These findings suggest targeted safety solutions including better roadway lighting and guardrails to reduce overturning-induced ejections, stricter rural seatbelt enforcement, and enhanced vehicle restraint systems to limit unintentional occupant movement.

10 April 2026 in Transportmetrica A: Transport Science
Verified
Journal Article
Closed access
MSMichael Starewich
Michael Starewich
SBSwastika Barua
Swastika Barua
ATAnannya Ghosh Tusti
Anannya Ghosh Tusti
SJSyed Aaqib Javed
Syed Aaqib Javed
SPSazzad Bin Bashar Polock
Sazzad Bin Bashar Polock
SDSubasish Das
Subasish Das

9 March 2026 in Accident Analysis & Prevention
Verified
Journal Article
Closed access
TCTausif Islam Chowdhury
Tausif Islam Chowdhury
MSMichael Starewich
Michael Starewich
SBSwastika Barua
Swastika Barua
ATAnannya Ghosh Tusti
Anannya Ghosh Tusti
SDSubasish Das
Subasish Das

17 December 2025 in Cities
Verified
Journal Article
Closed access
SJSyed Aaqib Javed
Syed Aaqib Javed
SBSwastika Barua
Swastika Barua
ATAnannya Ghosh Tusti
Anannya Ghosh Tusti
SPSazzad Bin Bashar Polock
Sazzad Bin Bashar Polock
TCTausif Islam Chowdhury
Tausif Islam Chowdhury
SDSubasish Das
Subasish Das