Senthooran Rajamanoharan
Scholar

Senthooran Rajamanoharan

Google Scholar ID: W4JthzgAAAAJ
Google DeepMind
Mechanistic InterpretabilityMachine Learning
Google Scholar↗
Citations & Impact
All-time
Citations
709
 
H-index
11
 
i10-index
11
 
Publications
20
 
Co-authors
18
list available
Contact
No contact links provided.
Publications
13 items
Emergent Misalignment is Easy, Narrow Misalignment is Hard
2026
Cited
0
Thought Branches: Interpreting LLM Reasoning Requires Resampling
2025
Cited
0
Eliciting Secret Knowledge from Language Models
2025
Cited
0
Steering Out-of-Distribution Generalization with Concept Ablation Fine-Tuning
2025
Cited
0
When Chain of Thought is Necessary, Language Models Struggle to Evade Monitors
2025
Cited
0
Dense SAE Latents Are Features, Not Bugs
2025
Cited
0
Convergent Linear Representations of Emergent Misalignment
2025
Cited
1
Model Organisms for Emergent Misalignment
2025
Cited
0
Resume (English only)
Co-authors
18 total
Neel Nanda
Neel Nanda
Mechanistic Interpretability Team Lead, Google DeepMind
Arthur Conmy
Arthur Conmy
Google DeepMind
Co-author 3
Co-author 3
Vikrant Varma
Vikrant Varma
DeepMind
Co-author 5
Co-author 5
Rohin Shah
Rohin Shah
Research Scientist, Google DeepMind
Co-author 7
Co-author 7
Brian Williams
Brian Williams
Professor of Aeronautics and Astronautics, Massachusetts Institute of Technology