Vikrant Varma

3 papers · 2024–2025 · 3 venues · across top CS/AI venues

Authors are matched by name, so researchers who share a name ("Yang Liu", "Wei Wang") appear as a single profile and their paper counts, venues and achievements are combined. How identity is handled.

Achievements

Jump to papers ↓

🌍 Conference Polyglot · Bronze 🌉 Interdisciplinary Bridge · Bronze 🐣 Hot Topic Early Bird

Venues

EMNLP (1) ICML (1) NeurIPS (1)

Top co-authors

Rohin Shah (3) Senthooran Rajamanoharan (2) Neel Nanda (2) Arthur Conmy (2) János Kramár (2) Tom Lieberum (2) Lewis Smith (2) Caleb Biddulph (1) David Lindner (1) Ian Goodfellow (1)

Research topics

Artificial Intelligence (3) Machine Learning (3) Core AI (3) Deep Learning (2) Architectures (2) Reinforcement Learning (1) Core Methods (1) Learning Types (1) Methods (1) Optimization & Theory (1)

Keywords

sparse autoencoder (2) feature learning (1) neural network interpretability (1) model analysis (1) latent representation (1) language model (1) feature decomposition (1) interpretable feature (1) large language model (1) activation decomposition (1) multi-step reward hacking (1) myopic optimization (1) non-myopic approval (1) reinforcement learning (1) oversight (1)

Papers

MONA: Myopic Optimization with Non-myopic Approval Can Mitigate Multi-step Reward Hacking ICML 2025 Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 EMNLP 2024 Improving Sparse Decomposition of Language Model Activations with Gated Sparse Autoencoders NeurIPS 2024