REVIEW 5 cited by
Gradients are Not All You Need
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
read the original abstract
Differentiable programming techniques are widely used in the community and are responsible for the machine learning renaissance of the past several decades. While these methods are powerful, they have limits. In this short report, we discuss a common chaos based failure mode which appears in a variety of differentiable circumstances, ranging from recurrent neural networks and numerical physics simulation to training learned optimizers. We trace this failure to the spectrum of the Jacobian of the system under study, and provide criteria for when a practitioner might expect this failure to spoil their differentiation based optimization algorithms.
Forward citations
Cited by 5 Pith papers
-
Open-DiffLoco: Open-Source Differentiable Learning for Deployable Blind Quadruped Locomotion
Differentiable simulation with SHAC and a simplified reward set trains deployable blind quadruped policies that transfer to a real Unitree Go2, plus a new critic-Jacobian supervision method (JAVE).
-
Efficient Domain-Adaptive Policy Learning via Kernel Representation with Application to Quadrotor Control under Non-Stationary Disturbances
A quadrotor policy trained in 50 seconds of differentiable simulation adapts online by updating both random-Fourier-feature coefficients and bandwidth, improving tracking over DATT/MPC baselines in simulation and on C...
-
RoboSSM: Scalable In-context Imitation Learning via State-Space Models
RoboSSM shows that a state-space model backbone can extend in-context imitation learning to prompts much longer than those seen in training, where a Transformer-based baseline degrades.
-
How Should We Meta-Learn Reinforcement Learning Algorithms?
A systematic comparison of black-box evolution, neural and symbolic distillation, and LLM-based proposal for meta-learning RL algorithms yields practical recommendations: warm-started LLM proposal is sample-efficient,...
-
First Order Model-Based RL through Decoupled Backpropagation
By computing gradients through a learned dynamics model while unrolling trajectories in the real simulator, DMO achieves SHAC-level sample efficiency with standard simulators and deploys on a real quadruped.
Discussion (0). Sign in to comment.