1 code implementation • 6 Mar 2024 • Shannon Zejiang Shen, Hunter Lang, Bailin Wang, Yoon Kim, David Sontag
We propose a method to teach multiple large language models (LLM) to collaborate by interleaving their generations at the token level.
1 code implementation • 15 Jan 2023 • Hussein Mozannar, Hunter Lang, Dennis Wei, Prasanna Sattigeri, Subhro Das, David Sontag
We show that prior approaches can fail to find a human-AI system with low misclassification error even when there exists a linear classifier and rejector that have zero error (the realizable setting).
1 code implementation • 19 Oct 2022 • Stefan Hegselmann, Alejandro Buendia, Hunter Lang, Monica Agrawal, Xiaoyi Jiang, David Sontag
We study the application of large language models to zero-shot and few-shot classification of tabular data.
1 code implementation • 6 Jun 2022 • Hunter Lang, Aravindan Vijayaraghavan, David Sontag
Subset selection applies to any label model and classifier and is very simple to plug in to existing weak supervision pipelines, requiring just a few lines of code.
no code implementations • 25 May 2022 • Monica Agrawal, Stefan Hegselmann, Hunter Lang, Yoon Kim, David Sontag
A long-running goal of the clinical NLP community is the extraction of important variables trapped in clinical notes.
1 code implementation • 2 Feb 2022 • Hunter Lang, Monica Agrawal, Yoon Kim, David Sontag
We demonstrate that co-training (Blum & Mitchell, 1998) can improve the performance of prompt-based learning by using unlabeled data.
no code implementations • 4 Nov 2021 • Monica Agrawal, Hunter Lang, Michael Offin, Lior Gazit, David Sontag
Label-scarce, high-dimensional domains such as healthcare present a challenge for modern machine learning techniques.
no code implementations • 27 Jul 2021 • Hoifung Poon, Hai Wang, Hunter Lang
We first present deep probabilistic logic(DPL), which offers a unifying framework for task-specific self-supervision by composing probabilistic logic with deep learning.
no code implementations • 26 Feb 2021 • Hunter Lang, Aravind Reddy, David Sontag, Aravindan Vijayaraghavan
Several works have shown that perturbation stable instances of the MAP inference problem in Potts models can be solved exactly using a natural linear programming (LP) relaxation.
no code implementations • 23 Dec 2020 • Hunter Lang, Hoifung Poon
Labeling training examples at scale is a perennial challenge in machine learning.
no code implementations • 7 Nov 2020 • Hunter Lang, David Sontag, Aravindan Vijayaraghavan
On "real-world" instances, MAP assignments of small perturbations of the problem should be very similar to the MAP assignment(s) of the original problem instance.
1 code implementation • 25 Feb 2020 • Pengchuan Zhang, Hunter Lang, Qiang Liu, Lin Xiao
We propose a statistical adaptive procedure called SALSA for automatically scheduling the learning rate (step size) in stochastic gradient methods.
1 code implementation • NeurIPS 2019 • Igor Gitman, Hunter Lang, Pengchuan Zhang, Lin Xiao
The use of momentum in stochastic gradient methods has become a widespread practice in machine learning.
no code implementations • 25 Sep 2019 • Pengchuan Zhang, Hunter Lang, Qiang Liu, Lin Xiao
We investigate statistical methods for automatically scheduling the learning rate (step size) in stochastic optimization.
no code implementations • NeurIPS 2019 • Hunter Lang, Pengchuan Zhang, Lin Xiao
Despite the development of numerous adaptive optimizers, tuning the learning rate of stochastic gradient methods remains a major roadblock to obtaining good practical performance in machine learning.
no code implementations • 12 Oct 2018 • Hunter Lang, David Sontag, Aravindan Vijayaraghavan
The simplest stability condition assumes that the MAP solution does not change at all when some of the pairwise potentials are (adversarially) perturbed.
no code implementations • 6 Nov 2017 • Hunter Lang, David Sontag, Aravindan Vijayaraghavan
Approximate algorithms for structured prediction problems---such as LP relaxations and the popular alpha-expansion algorithm (Boykov et al. 2001)---typically far exceed their theoretical performance guarantees on real-world instances.