6 papers
GradInf: Gradient Estimation as Probabilistic Inference
Gaurav Arya, Mathieu Huot, Moritz Schauer +2
Gradient estimation -- the task of computing the gradient of the expected value of a probabilistic program -- has diverse applications in scientific computing, but is notoriously d…
Incremental Computation for Efficient Programmable Inference in Probabilistic Programs
Fabian Zaiser, Jack Czenszak, Martin C. Rinard +2
Inference in probabilistic programs generally requires evaluating many possible program executions to find those of high posterior density. To scale inference to large datasets, it…
Fast Controlled Generation from Language Models with Adaptive Weighted Rejection Sampling
Benjamin Lipkin, Benjamin LeBrun, Jacob Hoover Vigly +9
The dominant approach to generating from language models subject to some constraint is locally constrained decoding (LCD), incrementally sampling tokens at each time step such that…
Self-Steering Language Models
Gabriel Grand, Joshua B. Tenenbaum, Vikash K. Mansinghka +2
While test-time reasoning enables language models (LMs) to tackle complex tasks, searching or planning in natural language can be slow, costly, and error-prone. But even when LMs s…
Modeling Open-World Cognition as On-Demand Synthesis of Probabilistic Models
Lionel Wong, Katherine M. Collins, Lance Ying +8
When faced with novel situations, people are able to marshal relevant considerations from a wide range of background knowledge and put these to use in inferences and predictions. W…
Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlo
João Loula, Benjamin LeBrun, Li Du +12
A wide range of LM applications require generating text that conforms to syntactic or semantic constraints. Imposing such constraints can be naturally framed as probabilistic condi…