proomt

Search

Search posts, papers, and topics

All posts

Hugging Face Daily PapersSebastián A. Cruz Romero1 min readpaperadvanced

Retention-Constrained Post-Training Quantization of Cellpose-SAM for Stem Cell Microscopy

Read on arxiv.org

Summary

The paper proposes a retention‑constrained post‑training quantization benchmark for Cellpose‑SAM, showing that weight‑only W8A16 and a mixed W4/W8 scheme keep instance F1 scores while cutting model size 6.76×, whereas ternary quantization fails on most images.

  • Retention is defined as the 95% bootstrap interval of mean F1 change staying above -0.02 for each imaging modality.
  • Weight‑only W8A16 preserves instance F1 across nuclei, fluorescence, and iPSC images with no failures.
  • A mixed W4/W8 scheme with four INT8 exceptions achieves 6.76× storage reduction without catastrophic errors on a 176‑field test set.
  • Ternary weight‑only quantization compresses 12.08× but catastrophically fails on 169 of 176 fields.

ML engineers building edge‑deployed microscopy pipelines need a rigorous, auditable way to compress foundation models without sacrificing diagnostic reliability.

8/10
computer visionml infraquantizationmedical imagingmodel compression

Related reading

  2. Training-Adaptive Convolutional Sparse Coding via Information Bottleneck for Robust Visual Representation

    This paper introduces a training-adaptive Convolutional Sparse Coding (CSC) framework where the sparsity coefficient is learned end-to-end via FISTA unfolding. It uses an information bottleneck perspective to balance representation compression and content preservation, showing improved robustness to input perturbations on CIFAR and ImageNet.

    Hugging Face Daily Papersarxiv.org1 minpaper

  4. Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization Problem

    The authors cast transformer block removal as a constrained binary optimization problem equivalent to an Ising glass, using a Hessian‑derived energy as a proxy for downstream quality. Solving the resulting QUBO with classical or quantum‑inspired solvers yields up to 23 MMLU points improvement over prior block‑removal baselines at 50 % depth compression.

    Hugging Facehuggingface.co8 min

  5. Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint

    Ternary Bonsai 2 27B is a new multimodal LLM based on Qwen3.8 27B, utilizing ternary weights for a 5.9GB model footprint. It achieves over 9x compression while retaining 98.2% of the full-precision model's aggregate benchmark performance across various tasks. This enables highly capable 27B-class models to run efficiently on local devices.

    Hacker News front pageprismml.com5 minreleaseHN579198lobste.rs12

  6. VC-Attention: Value Smoothing and Softmax Casting for Low-bit Attention

    VC-Attention introduces a training‑free low‑bit attention pipeline for diffusion transformers. It smooths value tensors via lightweight online clustering (V‑Smooth) and quantizes only the residual after subtracting block means, restoring the mean from the softmax row sum. It also replaces the FP32 softmax exponential with a fused FP8 cast (ExpCast‑FP8) that maps log‑scores directly to E4M3 probab…

    Hugging Face Daily Papersarxiv.org1 minpaper