REVIEW 8 cited by
Neural Contextual Bandits with Deep Representation and Shallow Exploration
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
abstract
We study a general class of contextual bandits, where each context-action pair is associated with a raw feature vector, but the reward generating function is unknown. We propose a novel learning algorithm that transforms the raw feature vector using the last hidden layer of a deep ReLU neural network (deep representation learning), and uses an upper confidence bound (UCB) approach to explore in the last linear layer (shallow exploration). We prove that under standard assumptions, our proposed algorithm achieves $\tilde{O}(\sqrt{T})$ finite-time regret, where $T$ is the learning time horizon. Compared with existing neural contextual bandit algorithms, our approach is computationally much more efficient since it only needs to explore in the last layer of the deep neural network.
Forward citations
Cited by 8 Pith papers
-
PinEqualizer: Full Funnel Content Exploration and Debiasing System at Pinterest
A full-funnel exploration and debiasing system deployed at Pinterest is reported to increase fresh-content impressions by ~350% and lift user-engagement and content-provider metrics.
-
Neural Variance-aware Dueling Bandits with Deep Representation and Shallow Exploration
Variance-aware neural dueling bandit algorithms achieve sublinear regret of order O(d sqrt(sum sigma_t^2) + sqrt(d T)) for wide networks on nonlinear utilities.
-
Online Clustering of Dueling Bandits
COLDB and CONDB are the first algorithms to combine online user clustering with dueling (preference) bandits, with regret bounds that improve as users are grouped into fewer clusters.
-
Contextual Bandit Optimization with Pre-Trained Neural Networks
E2TC achieves Õ((KT)^{4/5}) regret for stochastic contextual bandits with pre-trained neural representations and locally convex risk, without requiring extremely wide networks.
-
Neural Logistic Bandits
For neural logistic bandits, the paper claims regret bounds of \tilde O(\tilde d \sqrt{\kappa T}) and \tilde O(\tilde d \sqrt{T/\kappa^*}), improving on the previous \tilde O(\kappa \tilde d \sqrt{T}).
-
Neural Contextual Bandits Under Delayed Feedback Constraints
Under sub-exponential reward delays, Delayed NeuralUCB achieves regret O(d~√T logT + d~^{3/2}D+ log^{3/2}T), with D+ depending on the expected delay.
-
Epinet for Content Cold Start
An epinet-based Thompson sampling module improved cold-start video recommendations in a live Facebook Reels A/B test, but the experiment does not isolate the uncertainty mechanism.
-
In-Domain African Languages Translation Using LLMs and Multi-armed Bandits
Bandit-based model selection matches or slightly improves on the best single NMT system for in-domain English-to-African translation, but the claimed high-confidence statistical support is absent.
Discussion (0). Continue with ORCID to comment.