Dipesh Mahato

Papers on Lattice

Total citations

Topics

h-index

Research focus

Constitutional AI & AI Ethics (1)Eval Frameworks & Benchmarks (1)Interpretability & Mechanistic Interp (1)

Frequent co-authors

Deepanshu Mody (1)Deepanshu Mody (1)Samarth Agarwal (1)Samar Agarwal (1)Utkarsh Mittal (1)Utkarsh Mittal (1)Dipesh Mahato (1)

Papers (1)

Jul 28, 2026

Deepanshu Mody +7Jul 28, 2026

Minimizing Targeted Activations: Input-Only Suppression of Evaluation-Awareness Latents in Large Language Models

Suppressing evaluation-aware latents in LLMs reveals that activation control does not equate to behavioral influence, challenging assumptions about model safety during evaluations.

Deepanshu Mody, Deepanshu Mody, Samarth Agarwal +5

Constitutional AI & AI Ethics Eval Frameworks & Benchmarks Interpretability & Mechanistic Interp

Dipesh Mahato

Research focus

Frequent co-authors

Papers (1)

Search

Dipesh Mahato

Research focus

Frequent co-authors

Papers (1)