exponential moving average
PulseAugur coverage of exponential moving average — every cluster mentioning exponential moving average across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New research identifies 'catastrophic overfitting' in AI training and proposes a fix
Researchers have identified a phenomenon called catastrophic overfitting (CO) in fast adversarial training (FAT) of neural networks. CO occurs when robustness to multi-step attacks rapidly degrades, even as the model co…
-
New technique enhances few-step video generation with representation forcing
Researchers have introduced Self Flow Forcing (SF$^2$), a novel method to improve few-step video generation by incorporating representation learning into distribution matching distillation. This approach addresses limit…
-
New method distinguishes telemetry data errors with 96% accuracy
Researchers have developed a new method to distinguish between process changes, instrument faults, and data corruption in streaming telemetry data. This approach uses a combination of online detectors, including a Kalma…
-
TecoPrompt enhances vision-language models with temporal-conservative prompt learning
Researchers have developed TecoPrompt, a novel framework for robust prompt learning in vision-language models, particularly effective under noisy supervision. This method utilizes optimal transport (OT) pseudo-labeling …
-
New benchmark reveals AI models struggle to predict research trends
Researchers have developed a new benchmark called Research Attention Prediction (RAP) to evaluate how well large language models can track shifts in research attention within the AI/ML field. The benchmark, covering 278…
-
New research explores advanced routing techniques for Mixture-of-Experts models · 4 sources tracked
Researchers are exploring new methods to improve the performance and specialization of Mixture-of-Experts (MoE) models. One approach focuses on aligning the geometric structures of routing states across different layers…
-
New CSWL framework enhances data-free knowledge distillation via frequency domain analysis
Researchers have developed a new framework called CSWL to improve data-free knowledge distillation (DFKD) by addressing issues in the frequency domain. Existing DFKD methods often rely too heavily on teacher model prefe…
-
EMASAM offers efficient, stable alternative to SAM for model generalization
Researchers have introduced EMASAM, a new optimization technique designed to improve model generalization while reducing computational cost. Unlike traditional Sharpness-Aware Minimization (SAM), EMASAM bypasses the nee…
-
New CausalGate framework enhances transformer efficiency by pruning modules
Researchers have developed CausalGate, a new framework designed to make transformer inference more efficient. Unlike previous methods that relied on observational heuristics, CausalGate uses an intervention-guided appro…
-
New SPOFA framework stabilizes heterogeneous knowledge distillation
Researchers have developed SPOFA, a new framework designed to stabilize heterogeneous knowledge distillation (HKD). HKD aims to transfer knowledge between different model architectures, such as Transformers and CNNs, bu…