Dip Roy

Identifiers

No identifiers captured yet.

Papers (4)

1. Posterior-Calibrated Causal Circuits in Variational Autoencoders: Why Image-Domain Interpretability Fails on Tabular Data cs.LG · 2026 · author #1
2. Detection Without Correction: A Robust Asymmetry in Activation-Based Hallucination Probing cs.CL · 2026 · author #1
3. Fundamental Limits of Neural Network Sparsification: Evidence from Catastrophic Interpretability Collapse cs.LG · 2026 · author #1
4. A Multi-Level Causal Intervention Framework for Mechanistic Interpretability in Variational Autoencoders cs.LG · 2025 · author #1

Mentions

No mention provenance yet.

Frequent Coauthors

• Rajiv Misra 4 shared papers
• Sanjay Kumar Singh 4 shared papers
• Anisha Roy 3 shared papers