CAQ-ZO aligns ZO query stencils to compander grids, eliminating query-time residual error and improving NF4 fine-tuning performance on Qwen and Llama models compared to standard quantized baselines.
https://arxiv.org/abs/2406.06385
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
SplitQ improves low-bit PTQ for VLMs by isolating modality-specific outlier channels via MOCD and applying dual-branch adaptive calibration via ACC, outperforming prior methods on six datasets across W4A8 to W3A2 settings.
Cool-chic 5.0 achieves 11% BD-rate savings over VVC with a decoder 250× less complex than modern autoencoders and 10× fewer encoding iterations than previous overfitted codecs.
Quantization of VLMs improves multiple reliability metrics beyond accuracy by damping high-rank spectral components and promoting reliance on robust low-rank features.
GNMR is a gradient-norm-based controller that maps local stability signals to budgeted recovery actions to stabilize low-precision LLM training while preserving quality.
Factorial experiments with over 1300 runs falsify the hypothesis that INT6 QAT needs a different LR schedule from higher precision and identify a 50M-parameter boundary for INT4 schedule sensitivity.
Quant.npu provides a fully static quantization pipeline for on-device LLMs on NPUs by combining rotation matrices, bit-width-aware initialization, two-stage selective optimization, and adaptive mixed precision.
citing papers explorer
-
Compander-Aligned Query Geometry for Quantized Zeroth-Order Optimization
CAQ-ZO aligns ZO query stencils to compander grids, eliminating query-time residual error and improving NF4 fine-tuning performance on Qwen and Llama models compared to standard quantized baselines.
-
Breaking Modality Heterogeneity in Low-Bit Quantization for Large Vision-Language Models
SplitQ improves low-bit PTQ for VLMs by isolating modality-specific outlier channels via MOCD and applying dual-branch adaptive calibration via ACC, outperforming prior methods on six datasets across W4A8 to W3A2 settings.
-
Cool-chic 5.0: Faster Encoding and Inter-Feature Entropy Modeling for Overfitted Image Compression
Cool-chic 5.0 achieves 11% BD-rate savings over VVC with a decoder 250× less complex than modern autoencoders and 10× fewer encoding iterations than previous overfitted codecs.
-
Less Precise Can Be More Reliable: A Systematic Evaluation of Quantization's Impact on VLMs Beyond Accuracy
Quantization of VLMs improves multiple reliability metrics beyond accuracy by damping high-rank spectral components and promoting reliance on robust low-rank features.
-
GNMR: Runtime Stability Control for Low-Precision Large Language Model Training
GNMR is a gradient-norm-based controller that maps local stability signals to budgeted recovery actions to stabilize low-precision LLM training while preserving quality.
-
Mapping the Schedule x Bit-Width Boundary in Sub-100M Quantisation-Aware Training
Factorial experiments with over 1300 runs falsify the hypothesis that INT6 QAT needs a different LR schedule from higher precision and identify a 50M-parameter boundary for INT4 schedule sensitivity.
-
Quant.npu: Enabling Efficient Mobile NPU Inference for on-device LLMs via Fully Static Quantization
Quant.npu provides a fully static quantization pipeline for on-device LLMs on NPUs by combining rotation matrices, bit-width-aware initialization, two-stage selective optimization, and adaptive mixed precision.