Pith. sign in

REVIEW 13 cited by

Maintaining Plasticity in Continual Learning via Regenerative Regularization

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

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 2308.11958 v3 pith:VGJRTLVI submitted 2023-08-23 cs.LG cs.AI

Saurabh Kumar , Henrik Marklund , Benjamin Van Roy This is my paper
classification cs.LGcs.AI
keywords plasticityparameterscontinualinitlearningregularizationtowardadapt
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

In continual learning, plasticity refers to the ability of an agent to quickly adapt to new information. Neural networks are known to lose plasticity when processing non-stationary data streams. In this paper, we propose L2 Init, a simple approach for maintaining plasticity by incorporating in the loss function L2 regularization toward initial parameters. This is very similar to standard L2 regularization (L2), the only difference being that L2 regularizes toward the origin. L2 Init is simple to implement and requires selecting only a single hyper-parameter. The motivation for this method is the same as that of methods that reset neurons or parameter values. Intuitively, when recent losses are insensitive to particular parameters, these parameters should drift toward their initial values. This prepares parameters to adapt quickly to new tasks. On problems representative of different types of nonstationarity in continual supervised learning, we demonstrate that L2 Init most consistently mitigates plasticity loss compared to previously proposed approaches.

Discussion (0). Sign in to comment.

Forward citations

Cited by 13 Pith papers

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

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

    cs.LG 2026-04 unverdicted novelty 7.0 of 10

    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.

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

    cs.LG 2026-04 conditional novelty 6.5 of 10

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

  3. Local Redundancy: An Information-Theoretic Measure of Plasticity from Synthetic Memorization

    cs.LG 2026-07 conditional novelty 6.0 of 10

    An information-theoretic 'local redundancy' is lower-bounded, via an entropy-cancellation argument, by the expected squared gradient norm on synthetic probe data, and this proxy modestly out-predicts existing plastici...

  4. Preserving Plasticity in Continual Learning via Dynamical Isometry

    cs.LG 2026-06 unverdicted novelty 6.0 of 10

    Dynamical isometry (Jacobian singular values near 1) preserves plasticity in continual learning; an isometry-promoting regularizer and decoupled AdamO optimizer match or beat prior methods on supervised and RL benchmarks.

  5. Rotation-Preserving Supervised Fine-Tuning

    cs.LG 2026-05 unverdicted novelty 6.0 of 10

    RPSFT improves the in-domain versus out-of-domain performance trade-off during LLM supervised fine-tuning by penalizing rotations in pretrained singular subspaces as a proxy for loss-sensitive directions.

  6. Weight Decay Improves Language Model Plasticity

    cs.LG 2026-02 conditional novelty 6.0 of 10

    Pretrained models trained with larger weight decay fine-tune better on downstream tasks, so the best pretraining checkpoint by loss is not always the best starting point for later training.

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

    cs.LG 2025-09 unverdicted novelty 6.0 of 10

    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.

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

    cs.LG 2026-07 conditional novelty 5.5 of 10

    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.

  9. Continual-RL for Generalization in Autonomous Racing on the RoboRacer Platform

    cs.RO 2026-07 conditional novelty 5.0 of 10

    SAC plus Continual Backpropagation, trained only on real multi-track data, fine-tunes in ~15 minutes on an unseen lower-friction RoboRacer track and outperforms MAP and MPC.

  10. SFT Overtraining Predicts Rank Inversion via Entropy Collapse Under RLVR

    cs.LG 2026-06 unverdicted novelty 5.0 of 10

    SFT depth increases pre-RL pass@1 but can cause entropy collapse that inverts GRPO outcomes on Qwen models via reduced group advantage variance.

  11. On the Stability of Growth in Structural Plasticity

    cs.LG 2026-05 unverdicted novelty 5.0 of 10

    Newborn units in growing neural networks are forward-active but backward-starved, receiving weaker gradients than existing units and creating integration challenges that make growth less reliable than pruning in compl...

  12. On the Stability of Growth in Structural Plasticity

    cs.LG 2026-05 unverdicted novelty 5.0 of 10

    Growth during training inserts new units into a specialized trajectory, making them forward-active but backward-starved with weaker gradients than existing units.

  13. Activation Function Design Sustains Plasticity in Continual Learning

    cs.LG 2025-09 unverdicted novelty 5.0 of 10

    Smooth-Leaky and Randomized Smooth-Leaky activations mitigate loss of plasticity in continual learning by targeting negative-branch shape and saturation behavior.

Pith tools