96 citations · 359 across the 71 of their papers we have counts for
8 papers · 1 filter
Towards Continual Learning: Mitigating Forgetting With Subspace Tuning
Arnav Chavan, Kusha Sareen, Rishabh Tiwari +6
Large language models (LLMs) acquire new skills through fine-tuning, but it can destroy capabilities they already have: catastrophic forgetting remains a central obstacle to contin…
Convex-Concave Reinforcement Learning
Shripad V. Deshmukh, Yaswanth Chittepu, Dhawal Gupta +2
Policy learning drives many of the most consequential and heavily-invested applications of reinforcement learning today. Yet the core optimization problem it rests on (maximizing e…
Safe Inference-Time Alignment via Lagrangian Reward Augmentation
Yaswanth Chittepu, Ativ Joshi, Sohini Chintala +1
Inference-time alignment steers a frozen language model during decoding using auxiliary reward signals, avoiding the cost of repeated weight updates. However, existing inference-ti…
Hierarchical Experimentalist Agents
Abhranil Chandra, Sankaran Vaidyanathan, Utsav Dhanuka +2
Large language models (LLMs) are increasingly used to act in the real world and augment human decision-making, yet most systems rely on parametric knowledge acquired through imitat…
The Curse of Multiple Mediators: Hidden Interaction Effects in Activation Patching
Sankaran Vaidyanathan, David Arbour, Aaron Mueller +2
Activation patching is the primary tool in mechanistic interpretability. It attributes causal responsibility for a model behavior to each of its individual components by estimating…
Training ML Models with Predictable Failures
Will Schwarzer, Scott Niekum
Estimating how often an ML model will fail at deployment scale is central to pre-deployment safety assessment, but a feasible evaluation set is rarely large enough to observe the f…
Regularized Latent Dynamics Prediction is a Strong Baseline For Behavioral Foundation Models
Pranaya Jajoo, Harshit Sikchi, Siddhant Agarwal +3
Behavioral Foundation Models (BFMs) produce agents with the capability to adapt to any unknown reward or task. These methods, however, are only able to produce near-optimal policie…
Safe RLHF Beyond Expectation: Stochastic Dominance for Universal Spectral Risk Control
Yaswanth Chittepu, Ativ Joshi, Rajarshi Bhattacharjee +1
Safe Reinforcement Learning from Human Feedback (RLHF) typically enforces safety through expected cost constraints, but the expectation captures only a single statistic of the cost…