Pith. sign in

REVIEW 9 cited by

Continual Backprop: Stochastic Gradient Descent with Persistent Randomness

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

arxiv 2108.06325 v3 pith:V4J6HRH7 submitted 2021-08-13 cs.LG

Continual Backprop: Stochastic Gradient Descent with Persistent Randomness

classification cs.LG
keywords backpropcontinuallearningdescentgradientalgorithmcontinuallystochastic
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The Backprop algorithm for learning in neural networks utilizes two mechanisms: first, stochastic gradient descent and second, initialization with small random weights, where the latter is essential to the effectiveness of the former. We show that in continual learning setups, Backprop performs well initially, but over time its performance degrades. Stochastic gradient descent alone is insufficient to learn continually; the initial randomness enables only initial learning but not continual learning. To the best of our knowledge, ours is the first result showing this degradation in Backprop's ability to learn. To address this degradation in Backprop's plasticity, we propose an algorithm that continually injects random features alongside gradient descent using a new generate-and-test process. We call this the \textit{Continual Backprop} algorithm. We show that, unlike Backprop, Continual Backprop is able to continually adapt in both supervised and reinforcement learning (RL) problems. Continual Backprop has the same computational complexity as Backprop and can be seen as a natural extension of Backprop for continual learning.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. To Retain or to Adapt? Generalizing Continual Learning

    stat.ML 2026-07 accept novelty 7.5

    When retention induces positive stationary bias, historical knowledge becomes an optimization liability beyond a closed-form Critical Task Duration; Predictive CL with finite windows can beat both full Joint-Task and ...

  2. Beyond Single-Model Optimization: Preserving Plasticity in Continual Reinforcement Learning

    cs.LG 2026-04 unverdicted novelty 7.0

    TeLAPA maintains archives of behaviorally diverse yet competent policies aligned in a shared latent space to preserve plasticity and enable faster recovery after interference in continual reinforcement learning.

  3. Beyond Single-Model Optimization: Preserving Plasticity in Continual Reinforcement Learning

    cs.LG 2026-04 conditional novelty 6.5

    TeLAPA preserves behaviorally diverse policy neighborhoods in a shared latent space, improving MiniGrid continual RL transfer, revisit recovery, and retention over single-model preservation.

  4. Stagnant Neuron: Towards Understanding the Plasticity Loss in Multi-Agent Reinforcement Learning Value Factorization Methods

    cs.LG 2026-06 unverdicted novelty 6.0

    KNIFE targets stagnant neurons in MARL value factorization by replacing them with a composite of frozen, re-initialized, and compensating units to restore plasticity while preserving cooperation knowledge.

  5. Barriers for Learning in an Evolving World: Mathematical Understanding of Loss of Plasticity

    cs.LG 2025-09 unverdicted novelty 6.0

    This paper defines loss of plasticity via stable manifolds in parameter space and identifies frozen units and cloned-unit manifolds as the main mechanisms that trap gradient trajectories in non-stationary settings.

  6. Calibrated Partial Resets: Preventing Policy Collapse in Continual Reinforcement Learning

    cs.LG 2026-07 conditional novelty 5.5

    Utility-scaled partial neuron resets prevent policy collapse in long-horizon continual RL while matching or beating binary-reset and uniform-decay baselines on several benchmarks.

  7. Attribution-Based Neuron Utility for Plasticity Restoration in Deep Networks

    cs.LG 2026-05 unverdicted novelty 5.0

    GXD estimates the first-order functional cost of replacing a neuron via gradient attribution to make adaptive resets more reliable for preserving plasticity in continual learning.

  8. Investigating Action Encodings in Recurrent Neural Networks in Reinforcement Learning

    cs.LG 2026-05 unverdicted novelty 5.0

    The authors compare multiple methods for incorporating action information into RNN state updates for RL and report empirical results on illustrative domains.

  9. Task Switching Without Forgetting via Proximal Decoupling

    cs.LG 2026-04 unverdicted novelty 5.0

    Operator splitting separates task optimization from proximal stability enforcement to achieve forgetting-free continual learning with SOTA benchmark results.