Methods › General › Stochastic Optimization › Lookahead

Lookahead

21 papers tagged archive 2025-07-28

archive 2025-07-28 Description, source and code snippet are the archive's method entry.

Lookahead is a type of stochastic optimizer that iteratively updates two sets of weights: "fast" and "slow". Intuitively, the algorithm chooses a search direction by looking ahead at the sequence of fast weights generated by another optimizer.

Algorithm 1 Lookahead Optimizer

Require Initial parameters ϕ₀, objective function L

Require Synchronization period k, slow weights step size α, optimizer A

   for t=1, 2, …

     Synchronize parameters θ_(t,0) ϕₜ₋₁

     for i=1, 2, …, k

       sample minibatch of data d ∼𝒟

       θ_(t,i) θ_(t,i-1) + A(L, θ_(t,i-1), d)

     endfor

     Perform outer update ϕₜ ϕₜ₋₁ + α(θ_(t,k) - ϕₜ₋₁)

   endfor

   return parameters ϕ

Source: Lookahead Optimizer: k steps forward, 1 step backSee Code · jettify/pytorch-optimizer

Papers archive 2025-07-28

21 shown of 21, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.

Tasks archive 2025-07-28

20 shown of 27 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.

TaskPapers
Language Modeling2
Language Modelling2
Machine Translation2
Translation2
Chemical Process1
Code Completion1
Code Generation1
Deep Learning1
Diversity1
Domain Generalization1
EMG Gesture Recognition1
Electromyography (EMG)1
GSM8K1
Gesture Recognition1
Heuristic Search1
HumanEval1
Image Classification1
Instruction Following1
Interpretable Machine Learning1
Lifelong learning1

Usage over time archive 2025-07-28

Papers per year tagged with Lookahead: 2019 to 2025, peak 7 7 0 2019: 1 paper 2019 2020: 3 papers 2020 2021: 2 papers 2021 2022: 0 papers 2022 2023: 7 papers 2023 2024: 4 papers 2024 2025: 4 papers 2025
Papers per year the archive tags with this method, by the paper's archive date (21 dated). Bars are counts, not a trend claim.

Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).

Categories archive 2025-07-28

Stochastic Optimization

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections