10 papers
Mind the Delay: Latency-aware Query Routing for Heterogeneous LLM pools
Shivam Patel, Akaash R. Parthasarathy, Ankur Mallick +2
Large language model (LLM) query routers typically assign each query to a suitable model that balances response accuracy and monetary cost. Subsequently, separate system-level load…
Emergent and Subliminal Misalignment Through the Lens of Data-Mediated Transfer
Baris Askin, Muhammed Ustaomeroglu, Anupam Nayak +3
Fine-tuning LLMs on narrow harmful datasets can induce Emergent Misalignment (EM), where models exhibit misaligned behavior far beyond the fine-tuning distribution. We argue that e…
Not All Turns Are Equally Hard: Adaptive Thinking Budgets For Efficient Multi-Turn Reasoning in Agents
Neharika Jali, Anupam Nayak, Gauri Joshi
As LLM reasoning performance plateaus, improving inference-time compute efficiency is crucial to mitigate overthinking and long thinking traces even for simple queries. Prior appro…
Federate the Router: Learning Language Model Routers with Sparse and Decentralized Evaluations
Baris Askin, Shivam Patel, Anupam Nayak +4
Large language models (LLMs) are increasingly accessed as remotely hosted services by edge and enterprise clients that cannot run frontier models locally. Since models vary widely…
LOCUS: Low-Dimensional Model Embeddings for Efficient Model Exploration, Comparison, and Selection
Shivam Patel, William Cocke, Gauri Joshi
The rapidly growing ecosystem of Large Language Models (LLMs) makes it increasingly challenging to manage and utilize the vast and dynamic pool of models effectively. We propose LO…
Sample Complexity of Average-Reward Q-Learning: From Single-agent to Federated Reinforcement Learning
Yuchen Jiao, Jiin Woo, Gen Li +2
Average-reward reinforcement learning offers a principled framework for long-term decision-making by maximizing the mean reward per time step. Although Q-learning is a widely used…
ProxRouter: Proximity-Weighted LLM Query Routing for Improved Robustness to Outliers
Shivam Patel, Neharika Jali, Ankur Mallick +1
Large language model (LLM) query routers are critical to modern AI platforms as they seek to improve efficiency by assigning inference queries to accurate, yet low-cost models. Par…
Achieving Logarithmic Regret in KL-Regularized Zero-Sum Markov Games
Anupam Nayak, Tong Yang, Osman Yagan +2
Reverse Kullback-Leibler (KL) divergence-based regularization with respect to a fixed reference policy is widely used in modern reinforcement learning to preserve the desired trait…
Federated Offline Reinforcement Learning: Collaborative Single-Policy Coverage Suffices
Jiin Woo, Laixi Shi, Gauri Joshi +1
Offline reinforcement learning (RL), which seeks to learn an optimal policy using offline data, has garnered significant interest due to its potential in critical applications wher…
Efficient Reinforcement Learning for Routing Jobs in Heterogeneous Queueing Systems
Neharika Jali, Guannan Qu, Weina Wang +1
We consider the problem of efficiently routing jobs that arrive into a central queue to a system of heterogeneous servers. Unlike homogeneous systems, a threshold policy, that rout…