SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.08085 · NeurIPS · 2021

Natural continual learning: success is a journey, not (just) a destination

Alberto Bernacchia, Ta-Chu Kao, Kristopher Jensen, Gido Van De Ven, Guillaume Hennequin

arXiv · PDF · Open in the Atlas

Code that ran

1 of 5 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 0 failures of our instrument; 4 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2106.08085.json · badge: /api/ran/2106.08085.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
tachukao/ncl canonical (classifier label)
classifier basis: classifier_2025
1 ran · 4 no recorded run
FunctionStatusWhere it lives
TaskSet Ran
ran on a synthesized input · constructed an object; no result was computed
tachukao/ncl/ncl/learners/laplace.py
code served (permissive licence) · get_code("3efa917468087d0d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, TaskSet: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
import numpy as np

torch.manual_seed(0)

# default_hp: a dictionary of hyperparameters
default_hp = {
    'learning_rate': 0.001,
    'batch_size': 2,
    'hidden_dim': 4,
    'num_layers': 2
}

# generate_trials: a function that generates trial data
def generate_trials(batch_size):
    """Generate synthetic trial data"""
    x = torch.randn(batch_size, 4)
    y = torch.randn(batch_size, 3)
    return x, y

# negloglik: a function that computes negative log likelihood
def negloglik(predictions, targets):
    """Compute negative log likelihood"""
    mse_loss = nn.MSELoss()
    return mse_loss(predictions, targets)

# fudge: a small numerical constant for numerical stability
fudge = 1e-8

# sample_target: a function that samples target outputs
def sample_target(batch_size):
    """Sample target outputs"""
    return torch.randn(batch_size, 3)

# performance: a function that evaluates performance
def performance(predictions, targets):
    """Compute performance metric (e.g., accuracy or correlation)"""
    mse = torch.mean((predictions - targets) ** 2)
    return -mse.item()  # negative MSE as performance (higher is better)

ARGS = {
    'default_hp': default_hp,
    'generate_trials': generate_trials,
    'negloglik': negloglik,
    'fudge': fudge,
    'sample_target': sample_target,
    'performance': performance
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type TaskSet — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a TaskSet output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
AbstractLearner Not yet run
unverified (harvested, no recorded run)
tachukao/ncl/ncl/learners/laplace.py
code served (permissive licence) · get_code("4806687f38714cd6")
How it ran

no invocation recorded.

KFAC Not yet run
unverified (harvested, no recorded run)
tachukao/ncl/ncl/learners/laplace.py
code served (permissive licence) · get_code("7cdf217b133bf6fc")
How it ran

no invocation recorded.

Laplace Not yet run
unverified (harvested, no recorded run)
tachukao/ncl/ncl/learners/laplace.py
code served (permissive licence) · get_code("6f73b6a8a0817470")
How it ran

no invocation recorded.

NCL Not yet run
unverified (harvested, no recorded run)
tachukao/ncl/ncl/learners/laplace.py
code served (permissive licence) · get_code("a44c7c07b4f6054e")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 5 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Biological agents are known to learn many different tasks over the course of their lives, and to be able to revisit previous tasks and behaviors with little to no loss in performance. In contrast, artificial agents are prone to 'catastrophic forgetting' whereby performance on previous tasks deteriorates rapidly as new ones are acquired. This shortcoming has recently been addressed using methods that encourage parameters to stay close to those used for previous tasks. This can be done by (i) using specific parameter regularizers that map out suitable destinations in parameter space, or (ii) guiding the optimization journey by projecting gradients into subspaces that do not interfere with previous tasks. However, these methods often exhibit subpar performance in both feedforward and recurrent neural networks, with recurrent networks being of interest to the study of neural dynamics supporting biological continual learning. In this work, we propose Natural Continual Learning (NCL), a new method that unifies weight regularization and projected gradient descent. NCL uses Bayesian weight regularization to encourage good performance on all tasks at convergence and combines this with gradient projection using the prior precision, which prevents catastrophic forgetting during optimization. Our method outperforms both standard weight regularization techniques and projection based approaches when applied to continual learning problems in feedforward and recurrent networks. Finally, the trained networks evolve task-specific dynamics that are strongly preserved as new tasks are learned, similar to experimental findings in biological circuits.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2106.08085")
get_code_for_paper("2106.08085")
have("2106.08085")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2106.08085.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2106.08085.svg)](https://syntology.ai/paper/2106.08085)

Connect an agent — have() is free.