Related papers: BEACON: Bayesian Optimal Stopping for Efficient LL…
Large Language Models (LLMs) often expend significant computational resources generating boilerplate responses, such as refusals, simple acknowledgements and casual greetings, which adds unnecessary cost and latency. To address this…
Adaptive prompt and program search makes LLM evaluation selection-sensitive. Once benchmark items are reused inside tuning, the observed winner's score need not estimate the fresh-data performance of the full tune-then-deploy procedure. We…
Reasoning in Large Language Models (LLMs) poses a challenge for oversight as many misaligned behaviors do not surface until reasoning concludes. To address this, we introduce Behavior Cue Reasoning for making LLM reasoning more controllable…
Large Language Models (LLMs) often exhibit knowledge disparities across languages. Encouraging LLMs to \textit{abstain} when faced with knowledge gaps is a promising strategy to reduce hallucinations in multilingual settings. Current…
The problem of constructing a dataset for MLIP development which gives the maximum quality in the minimum amount of compute time is complex, and can be approached in a number of ways. We introduce a ``Bayesian selection" approach for…
We study the problem of optimal sampling in an edge-based video analytics system (VAS), where sensor samples collected at a terminal device are offloaded to a back-end server that processes them and generates feedback for a user. Sampling…
We study Bayesian optimal control of a general class of smoothly parameterized Markov decision problems. Since computing the optimal control is computationally expensive, we design an algorithm that trades off performance for computational…
The problem of sequential anomaly detection is considered, where multiple data sources are monitored in real time and the goal is to identify the "anomalous" ones among them, when it is not possible to sample all sources at all times. A…
The efficient use of limited computational resources is an essential ingredient of intelligence. Selecting computations optimally according to rational metareasoning would achieve this, but this is computationally intractable. Inspired by…
The high simulation cost has been a bottleneck of practical analog/mixed-signal design automation. Many learning-based algorithms require thousands of simulated data points, which is impractical for expensive to simulate circuits. We…
Test-time scaling improves the inference performance of Large Language Models (LLMs) but also incurs substantial computational costs. Although recent studies have reduced token consumption through dynamic self-consistency, they remain…
Sequential estimation of a vector of linear regression coefficients is considered under both centralized and decentralized setups. In sequential estimation, the number of observations used for estimation is determined by the observed…
Off-policy deep reinforcement learning (RL) typically leverages replay buffers for reusing past experiences during learning. This can help improve sample efficiency when the collected data is informative and aligned with the learning…
Large language models (LLMs) have been proposed as alternatives to human experts for estimating unknown quantities with associated uncertainty, a process known as Bayesian elicitation. We test this by asking eleven LLMs to estimate…
We study the problem of estimating the source of a network cascade. The cascade starts from a single vertex at time 0 and spreads over time, but only a noisy version of the propagation is observable. The goal is then to design a stopping…
The challenge of balancing user relevance and content diversity in recommender systems is increasingly critical amid growing concerns about content homogeneity and reduced user engagement. In this work, we propose a novel framework that…
Preferential Bayesian optimization allows optimization of objectives that are either expensive or difficult to measure directly, by relying on a minimal number of comparative evaluations done by a human expert. Generating candidate…
We consider a small extent sensor network for event detection, in which nodes take samples periodically and then contend over a {\em random access network} to transmit their measurement packets to the fusion center. We consider two…
Bayesian model-based reinforcement learning is a formally elegant approach to learning optimal behaviour under model uncertainty, trading off exploration and exploitation in an ideal way. Unfortunately, finding the resulting Bayes-optimal…
In this thesis, we introduce Bayesian filtering as a principled framework for tackling diverse sequential machine learning problems, including online (continual) learning, prequential (one-step-ahead) forecasting, and contextual bandits. To…