Pith. sign in

REVIEW 25 cited by

Symbolic Discovery of Optimization Algorithms

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 2302.06675 v4 pith:RCEJQYLZ submitted 2023-02-13 cs.LG cs.AIcs.CLcs.CVcs.NE

Symbolic Discovery of Optimization Algorithms

classification cs.LG cs.AIcs.CLcs.CVcs.NE
keywords lionadamtextbftrainingoptimizationprogramsearchtextit
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

We present a method to formulate algorithm discovery as program search, and apply it to discover optimization algorithms for deep neural network training. We leverage efficient search techniques to explore an infinite and sparse program space. To bridge the large generalization gap between proxy and target tasks, we also introduce program selection and simplification strategies. Our method discovers a simple and effective optimization algorithm, $\textbf{Lion}$ ($\textit{Evo$\textbf{L}$ved S$\textbf{i}$gn M$\textbf{o}$me$\textbf{n}$tum}$). It is more memory-efficient than Adam as it only keeps track of the momentum. Different from adaptive optimizers, its update has the same magnitude for each parameter calculated through the sign operation. We compare Lion with widely used optimizers, such as Adam and Adafactor, for training a variety of models on different tasks. On image classification, Lion boosts the accuracy of ViT by up to 2% on ImageNet and saves up to 5x the pre-training compute on JFT. On vision-language contrastive learning, we achieve 88.3% $\textit{zero-shot}$ and 91.1% $\textit{fine-tuning}$ accuracy on ImageNet, surpassing the previous best results by 2% and 0.1%, respectively. On diffusion models, Lion outperforms Adam by achieving a better FID score and reducing the training compute by up to 2.3x. For autoregressive, masked language modeling, and fine-tuning, Lion exhibits a similar or better performance compared to Adam. Our analysis of Lion reveals that its performance gain grows with the training batch size. It also requires a smaller learning rate than Adam due to the larger norm of the update produced by the sign function. Additionally, we examine the limitations of Lion and identify scenarios where its improvements are small or not statistically significant. Lion is also successfully deployed in production systems such as Google search ads CTR model.

discussion (0)

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

Forward citations

Cited by 25 Pith papers

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

  1. Optimizer Memory Makes Shuffle Order a First-Order Source of Fine-Tuning Noise

    cs.LG 2026-06 unverdicted novelty 7.0

    Fixed-clock optimizer memory turns equal-multiset data shuffle order into an O(η) source of fine-tuning noise, larger than the O(η²) effect in memoryless cases, with a fit-free sizing method derived.

  2. Scalable Neural Decoders for Practical Fault-Tolerant Quantum Computation

    quant-ph 2026-04 unverdicted novelty 7.0

    Neural decoder for quantum LDPC codes achieves ~10^{-10} logical error at 0.1% physical error with 17x improvement and high throughput, enabling practical fault tolerance at modest code sizes.

  3. OmniMol: Transferring Particle Physics Knowledge to Molecular Dynamics with Point-Edge Transformers

    physics.chem-ph 2026-01 unverdicted novelty 7.0

    OmniMol transfers a billion-jet pre-trained PET foundation model from HEP to molecular dynamics via an interaction-matrix attention bias, delivering strong performance on the oMol dataset with minimal fine-tuning and ...

  4. On the Convergence of Muon and Beyond

    cs.LG 2025-09 unverdicted novelty 7.0

    Muon-MVR2 attains the optimal anytime convergence rate of ~O(T^{-1/3}) in stochastic non-convex settings under horizon-free schedules.

  5. Causal Optimizer Interaction Calculus: Hidden Geometric Relaxation and Identifiable Interventions

    cs.LG 2026-07 conditional novelty 6.0

    Under fixed innovation coupling, finite-horizon optimizers admit minimal pathwise realizations and incidence-identifiable Möbius effects, with a five-term readout transfer from hidden relaxation and a closed reduced-v...

  6. Improving Neural Network Training by Decoupling the Magnitude and Direction of Weight Vectors

    cs.LG 2026-06 unverdicted novelty 6.0

    MD Decoupling factorizes weights into fixed-norm directions and learnable per-row/column magnitudes updated at independent rates, improving Adam and Muon training stability and scale transfer without weight decay or warmup.

  7. Sparse Mamba Decoder for Quantum Error Correction: Efficient Defect-Centric Processing of Surface Code Syndromes

    quant-ph 2026-05 unverdicted novelty 6.0

    Sparse Mamba Decoder processes only active defects in surface code syndromes using a 13-feature representation and Mamba backbone for O(k) complexity, reporting speedups and accuracy gains over dense decoders.

  8. Sparse Mamba Decoder for Quantum Error Correction: Efficient Defect-Centric Processing of Surface Code Syndromes

    quant-ph 2026-05 conditional novelty 6.0

    Sparse Mamba Decoder uses defect-centric 13D features and Mamba backbone for O(k) complexity decoding of surface code syndromes, cutting latency and improving accuracy over dense neural and classical decoders.

  9. Anon: Extrapolating Adaptivity Beyond SGD and Adam

    cs.AI 2026-05 unverdicted novelty 6.0

    Anon optimizer uses tunable adaptivity and incremental delay update to achieve convergence guarantees and outperform existing methods on image classification, diffusion, and language modeling tasks.

  10. GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling

    cs.CL 2026-04 unverdicted novelty 6.0

    GSQ applies a Gumbel-Softmax relaxation to learn discrete grid assignments in scalar quantization, closing most of the accuracy gap to vector methods like QTIP on Llama-3.1 models at 2-3 bits while using only symmetri...

  11. GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling

    cs.CL 2026-04 unverdicted novelty 6.0

    GSQ uses Gumbel-Softmax to optimize scalar quantization grids for LLMs, closing most of the accuracy gap to vector methods like QTIP at 2-3 bits per parameter while using symmetric scalar grids compatible with existin...

  12. Diffusion-Based Point-Cloud Generation of Heavy-Ion Events

    hep-ph 2026-04 unverdicted novelty 6.0

    A two-stage score-driven diffusion model with Point-Edge Transformer generates realistic high-multiplicity heavy-ion events as point clouds.

  13. Explicit or Implicit? Encoding Physics at the Precision Frontier

    hep-ph 2026-03 conditional novelty 6.0

    On three precision classification tasks — reweighting-based unfolding, likelihood-ratio estimation, and weakly supervised anomaly detection — a Lorentz-equivariant transformer and a pretrained foundation model perform...

  14. Minimum Bayes Risk Decoding for Error Span Detection in Reference-Free Automatic Machine Translation Evaluation

    cs.CL 2025-12 conditional novelty 6.0

    Minimum Bayes Risk decoding with a soft F1 utility improves LLM error-span detection over standard MAP decoding on WMT24, and DPO distillation recovers the gains at greedy-search speed.

  15. The Falcon Series of Open Language Models

    cs.CL 2023-11 conditional novelty 6.0

    Falcon-180B is a 180B-parameter open decoder-only model trained on 3.5 trillion tokens that approaches PaLM-2-Large performance at lower cost and is released with dataset extracts.

  16. Token Geometry

    cs.LG 2026-07 conditional novelty 5.5

    Ember matches Adam on embedding/LM-head tables with O(V+D) state by factoring second moments and dropping first-moment momentum, and token trajectories look nearly 1D.

  17. Searching the Space of Feed-Forward Neural-Network Weight-Update Rules with Fixed Depth Symbolic Regression

    cs.LG 2026-07 conditional novelty 5.0

    Symbolic regression over fixed-depth expressions found neural-network weight-update rules that beat tuned standard optimizers on 25 of 30 small benchmark/NN setups, but evaluation used the same data for selection and ...

  18. Causal Optimizer Interaction Calculus: Hidden Geometric Relaxation and Identifiable Interventions

    cs.LG 2026-07 conditional novelty 5.0

    A modular calculus decomposes optimizer updates into geometric preconditioning plus structured nongeometric mechanisms, with a direction-expressivity theorem showing full SPD geometry captures exactly strict descent d...

  19. Token Geometry

    cs.LG 2026-07 unverdicted novelty 5.0

    Ember is a memory-efficient optimizer for token embeddings that exploits distinct gradient geometry and models token trajectories as 1D rays.

  20. Improving Neural Network Training by Decoupling the Magnitude and Direction of Weight Vectors

    cs.LG 2026-06 conditional novelty 5.0

    Splitting weight matrices into a fixed-norm direction and learnable per-row/column magnitudes improves LLM training over AdamW/Muon, removes weight decay and warmup, and transfers the optimal LR across width.

  21. Enhancing SignSGD: Small-Batch Convergence Analysis and a Hybrid Switching Strategy

    cs.LG 2026-04 unverdicted novelty 5.0

    SignSGD with pre-sign dithering and a calibrated hybrid switch to SGD achieves 92.18% accuracy on CIFAR-10 with ResNet-18, outperforming pure SGD and SignSGD, plus better results than Adam on CIFAR-100.

  22. DINOv2: Learning Robust Visual Features without Supervision

    cs.CV 2023-04 unverdicted novelty 5.0

    Pith review generated a malformed one-line summary.

  23. Xray-Visual Models: Scaling Vision models on Industry Scale Data

    cs.CV 2026-02 conditional novelty 4.0

    A 2-billion-parameter vision encoder trained on 15B+ image-text and billions of video-hashtag pairs reports SOTA ImageNet linear-probe, Kinetics, and retrieval numbers, but relies on proprietary data and has several v...

  24. HyBDM: Multi-Scale Hybrid Experts for Time Series Forecasting with Bidirectional Dependency Modeling

    cs.LG 2026-07 conditional novelty 3.0

    HyBDM combines a Mamba-style global-pattern expert with a local window transformer and a learned router to forecast multivariate time series, reporting state-of-the-art results on six benchmarks.

  25. Navigating LLM Valley: From AdamW to Memory-Efficient and Matrix-Based Optimizers

    cs.LG 2026-05 unverdicted novelty 3.0

    This survey organizes LLM optimizer literature into categories and argues the field is shifting toward rigorous, multi-factor comparisons of convergence, memory, stability, and complexity.