Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,657 papers · 148 categories

Trend · papers per month

73147220293 · Jun 202019922001200920172026
48 results for CLIP loss

Study on gradient clipping in SGD for high-dimensional problems.

problem Understanding and optimizing gradient clipping in high-dimensional machine learning models.
method Theoretical analysis of streaming SGD with gradient clipping in a least squares problem, focusing on large intrinsic dimensionality.
result Developed a deterministic equation to describe the loss evolution in clipped SGD, showing benefits under certain conditions.

We consider the problem of recovering a low-rank matrix from its clipped observations. Clipping is conceivable in many scientific areas that obstructs statistical analyses. On the other hand, matrix completion (MC) methods can recover a low-rank matrix from various information deficits by using the principle of low-ran…

2018-09-13abs ↗pdf ↗

Unified CLIP space manipulations improve GAN adaptation with a single target image.

problem Overfitting or underfitting in fine-tuning a pre-trained generator with a single target image.
method Two-step training strategy: latent optimization in CLIP space followed by generator fine-tuning with CLIP space consistency loss.
result Our model generates diverse outputs with the target texture and outperforms baseline models.

Existing approaches for training neural networks with user-level differential privacy (e.g., DP Federated Averaging) in federated learning (FL) settings involve bounding the contribution of each user's model update by clipping it to some constant value. However there is no good a priori setting of the clipping norm acr…

2019-05-09abs ↗pdf ↗

New method reduces privacy impact on model accuracy for underrepresented groups.

problem Privacy mechanisms disproportionately affect underrepresented groups in machine learning models.
method Proposes DPSGD-F, a modified DPSGD that adjusts group contributions based on clipping bias.
result DPSGD-F removes disparate impact of differential privacy on model accuracy for protected groups.

Artificial neural network training with stochastic gradient descent can be destabilized by "bad batches" with high losses. This is often problematic for training with small batch sizes, high order loss functions or unstably high learning rates. To stabilize learning, we have developed adaptive learning rate clipping (A…

2019-06-21abs ↗pdf ↗

New privacy bounds for DP-SGD's last iterate, even with cyclic sampling.

problem Privacy of the last iterate in DP-SGD with cyclic sampling.
method Established new RDP upper bounds for the last iterate under realistic assumptions.
result Privacy bounds for DP-SGD's last iterate with cyclic sampling and clipping, even for nonconvex losses.

PPO's gradients are heavy-tailed, affecting learning; a robust estimator improves performance.

problem Heavy-tailedness of PPO gradients causing learning issues.
method Characterized heavy-tailed gradients, identified likelihood ratios and advantages as sources, proposed GMOM as a robust estimator.
result GMOM improves PPO performance without clipping tricks.

A new method for image translation using disentangled style and content preservation.

problem Difficulty in maintaining original content during reverse diffusion in diffusion-based image translation.
method Disentangled style and content representation using intermediate keys from ViT model, CLIP loss, semantic divergence loss, and resampling strategy.
result Outperforms state-of-the-art models in text-guided and image-guided translation tasks.

Improved differentially private deep learning with group-wise clipping techniques.

problem Efficiency and privacy trade-offs in deep learning models.
method Group-wise clipping techniques (per-layer and per-device) to reduce compute time and memory overhead.
result Private learning with group-wise clipping achieves similar or better performance than non-private learning with less wall time.

Private statistics estimation faces a bias, accuracy, and privacy trilemma.

problem Balancing privacy, accuracy, and bias in statistical estimation.
method Use differential privacy (DP) for private statistics, but clip samples to control sensitivity and add noise for privacy, introducing bias.
result No algorithm can simultaneously have low bias, low error, and low privacy loss for arbitrary distributions.

Privacy-preserving SGD with heavy-tailed noise achieves differential privacy guarantees.

problem Privacy preservation in noisy SGD with heavy-tailed noise.
method Differential privacy guarantees for SGD with heavy-tailed noise.
result SGD with heavy-tailed perturbations achieves (0,O(1/n))(0, O(1/n))-DP.

New method improves fairness in DP learning by preventing excessive gradient suppression.

problem Disparate impact on model predictions for minority groups in DP learning.
method Bounded adaptive clipping to prevent excessive gradient suppression.
result Improves worst-class accuracy by over 10 percentage points compared to existing methods.

Paper investigates multimodal contrastive learning and incorporates unpaired data.

problem Improving feature learning ability of multimodal models under noisy data.
method Initiates investigation of nonlinear loss functions for multimodal contrastive learning, analyzes performance, proposes new loss incorporating unpaired data.
result MMCL can outperform unimodal contrastive learning and robustly handle noisy data.

EpiMer merges models by solving Fréchet mean on a Riemannian manifold.

problem Integrating knowledge from multiple models without retraining.
method EpiMer casts model merging as solving the Fréchet mean on a Riemannian manifold, restricting computation to a low-rank subspace.
result EpiMer outperforms flat-geometry methods on image classification tasks.

This paper introduces 'General Cyclical Training' for neural networks.

problem Improving training efficiency and performance of neural networks.
method Cyclical training phases with varying hyperparameters, batch sizes, loss functions, and data augmentation.
result Cyclical weight decay, softmax temperature, and gradient clipping enhance model accuracy.

This paper analyzes convergence of DP-SGD with adaptive quantile clipping.

problem Empirical success of adaptive clipping methods lacks theoretical understanding.
method Comprehensive convergence analysis of SGD with quantile clipping (QC-SGD).
result Establishes theoretical guarantees for DP-QC-SGD, revealing relationships between quantile selection, step size, and convergence.

Differentially private GANs improve image privacy without significant quality loss.

problem Anonymizing image data sets while maintaining image quality.
method Training GANs with differential privacy on MNIST, analyzing privacy-utility trade-offs and explaining optimization methods.
result An increasing privacy budget adds little to generated image quality, revealing a saturated training regime.

CLIP models robustness to spurious features is re-evaluated using a new dataset.

problem Existing robustness tests of CLIP models may not fully reflect their performance on spurious features.
method Crafted a new dataset (CounterAnimal) to reveal CLIP models' reliance on realistic spurious features.
result CLIP models are robust to spurious features learned from their training data, not ImageNet.

CLIP learns joint image-text representations for zero-shot learning.

problem Understanding and improving zero-shot transfer performance in CLIP.
method Formal study of transferrable representation learning and analysis of zero-shot transfer performance.
result Proposes a new CLIP-type approach that outperforms existing methods.

Paper introduces Latent-CLIP for efficient text-image comparison in latent space.

problem Efficiently compare text and images in latent space without costly decoding.
method Trains CLIP model in latent space, uses Latent-CLIP rewards for noise optimization, and guides generation away from harmful content.
result Latent-CLIP matches CLIP performance on text-image classification and harmful content detection.

This paper improves analysis of clipping algorithms for non-convex optimization.

problem Improving convergence analysis of clipping algorithms for non-convex optimization.
method Develops a general framework to study clipping algorithms, including momentum methods, and provides convergence analysis in deterministic and stochastic settings.
result Demonstrates that clipping methods can maintain efficiency even in highly non-smooth regions.

Enhances CLIP's similarity computation using PMI's linear structure.

problem CLIP's similarity computation misses the optimal linear structure of PMI.
method KME-CLIP, utilizing inner product in a reproducing kernel Hilbert space.
result KME-CLIP approximates PMI with arbitrary accuracy and outperforms CLIP.

This paper improves convergence guarantees for gradient clipping in deep learning.

problem Improving convergence guarantees for gradient clipping in deep learning models.
method Analyzes and provides precise convergence guarantees for arbitrary clipping thresholds.
result Shows tight convergence guarantees for clipped stochastic gradient descent.

Clip21 improves convergence of gradient-clipped methods in DP settings.

problem Gradient clipping introduces bias in distributed training, causing convergence issues.
method Clip21 designs an error feedback mechanism to mitigate bias in gradient-clipped methods.
result Clip21 converges at the same rate as distributed gradient descent, improving from previous $\mathcal{O}\left(\frac{1}{\sqrt{K}} ight)$ to $\mathcal{O}\left(\frac{1}{K} ight)$.

Paper proves convergence for private FL on non-Lipschitz convex objectives using normalization instead of clipping.

problem Lack of convergence results for differentially private federated learning with non-Lipschitz objectives.
method Developed a convergence result for private FL on smooth convex objectives without assuming Lipschitzness, using normalization instead of clipping.
result Normalization-based private FL algorithm converges better than clipping-based counterpart on smooth convex functions.

INF-clip optimizes heavy-tailed MAB problems with improved performance.

problem Optimizing multi-armed bandit problems with heavy-tailed rewards.
method INF-clip algorithm for adversarial and stochastic heavy-tailed MAB settings.
result INF-clip is optimal for linear and non-linear heavy-tailed stochastic MAB problems.

Paper explores whether gradient normalization can replace clipping for SGD in heavy-tailed noise.

problem Ensuring convergence of SGD in heavy-tailed noise.
method Revisits gradient clipping and normalization, proving their sufficiency and effectiveness.
result Gradient normalization alone is sufficient for nonconvex SGD convergence under smoothness assumptions.

Many continuous control tasks have bounded action spaces. When policy gradient methods are applied to such tasks, out-of-bound actions need to be clipped before execution, while policies are usually optimized as if the actions are not clipped. We propose a policy gradient estimator that exploits the knowledge of action…

2018-02-21abs ↗pdf ↗

We propose a novel actor-critic, model-free reinforcement learning algorithm which employs a Bayesian method of parameter space exploration to solve environments. A Gaussian process is used to learn the expected return of a policy given the policy's parameters. The system is trained by updating the parameters using gra…

2020-03-02abs ↗pdf ↗

TROLL improves RL for LLMs by replacing clipping with a trust region projection.

problem Clipping in RL for LLMs causes instability and suboptimal performance.
method TROLL uses a discrete differentiable trust region projection to replace clipping, balancing computational cost and effectiveness.
result TROLL consistently outperforms PPO-like clipping in training speed, stability, and final success rates.

We consider the problem of minimizing a sum of clipped convex functions; applications include clipped empirical risk minimization and clipped control. While the problem of minimizing the sum of clipped convex functions is NP-hard, we present some heuristics for approximately solving instances of these problems. These h…

2019-10-27abs ↗pdf ↗

Theory for RLHF generalization under reward shift and clipped KL.

problem Theoretical understanding of RLHF generalization, especially with reward shift and clipped KL.
method Developed generalization theory for RLHF, accounting for reward shift and clipped KL.
result Presented generalization bounds for RLHF, suggesting generalization error from sampling, reward shift, and KL clipping.

Improved Clipped-SGD achieves near-optimal heavy-tailed statistical estimation in streaming settings.

problem High-dimensional heavy-tailed statistical estimation in streaming with memory constraints.
method Stochastic convex optimization with Clipped-SGD, proving near-optimal sub-Gaussian statistical rates.
result Clipped-SGD achieves an error of Tr(Σ)+Tr(Σ)Σ2log(log(T)δ)T\sqrt{\frac{\mathsf{Tr}(Σ)+\sqrt{\mathsf{Tr}(Σ)\|Σ\|_2}\log(\frac{\log(T)}δ)}{T}} with probability 1δ1-δ.

New optimization method combines gradient clipping and non-Euclidean smoothness.

problem Improving optimization in non-Euclidean spaces for machine learning.
method Hybrid of steepest descent and conditional gradient, incorporating weight decay.
result Achieves optimal convergence rate and demonstrates effectiveness in deep learning.