Affinity Posters
Blog Track Session 7
David Dobre · Leo Schwinn · Claire Vernade · Charlie Gauthier · Fabian Pedregosa · Gauthier Gidel
Halle B
Live content is unavailable. Log in and register to view live content
Schedule
Fri 1:45 a.m. - 3:45 a.m.
|
Deep Equilibrium Models For Algorithmic Reasoning
(
Poster
#3
)
>
link
Poster Location: Halle B #3 In this blogpost we discuss the idea of teaching neural networks to reach fixed points when reasoning. Specifically, on the algorithmic reasoning benchmark CLRS the current neural networks are told the number of reasoning steps they need. While a quick fix is to add a termination network that predicts when to stop, a much more salient inductive bias is that the neural network shouldn't change it's answer any further once the answer is correct, i.e. it should reach a fixed point. This is supported by denotational semantics, which tells us that while loops that terminate are the minimum fixed points of a function. We implement this idea with the help of deep equilibrium models and discuss several hurdles one encounters along the way. We show on several algorithms from the CLRS benchmark the partial success of this approach and the difficulty in making it work robustly across all algorithms. |
Louis-Pascal Xhonneux · Yu He · Andreea Deac · Jian Tang · Gauthier Gidel 🔗 |
Fri 1:45 a.m. - 3:45 a.m.
|
A Critical Exploration of "Bayesian Model Selection, Marginal Likelihood, and Generalization in Neural Networks"
(
Poster
#2
)
>
link
Poster Location: Halle B #2 This blog post comprehensively reviews the ICML 2022 paper titled "Bayesian Model Selection, the Marginal Likelihood, and Generalization." The paper's central thesis, examining the log marginal likelihood (LML) and its variant, the conditional log marginal likelihood (CLML), in various machine learning settings, is thoroughly analyzed, and the post critically engages with the paper's methodologies and findings, particularly scrutinizing the CLML's applicability and effectiveness in deep learning scenarios. The review extends beyond summarization to challenge assumptions, compare with existing literature, and examine the evaluation. This deep dive aims to foster a better understanding of Bayesian methods in model evaluation, spotlighting both their strengths and limitations in the context of neural network generalization. |
Andreas Kirsch 🔗 |
Fri 1:45 a.m. - 3:45 a.m.
|
Fair Model-Based Reinforcement Learning Comparisons with Explicit and Consistent Update Frequency
(
Poster
#1
)
>
link
Poster Location: Halle B #1 Implicit update frequencies can introduce ambiguity in the interpretation of model-based reinforcement learning benchmarks, obscuring the real objective of the evaluation. While the update frequency can sometimes be optimized to improve performance, real-world applications often impose constraints, allowing updates only between deployments on the actual system. This blog post emphasizes the need for evaluations using consistent update frequencies across different algorithms to provide researchers and practitioners with clearer comparisons under realistic constraints. |
Albert Thomas · Abdelhakim Benechehab · Giuseppe Paolo · Balázs Kégl 🔗 |