#model compression

23 results
cs.AI2026

WIDE: Boosting Adaptive LLM Inference via Token-level Dynamic Width Pruning

Haozhe Hu, Hao Wu, Peiran Yin +3

WIDE introduces a token-level dynamic width pruning framework for large language model inference, allowing each token to selectively activate attention heads and feed‑forward chann…

#large language models#dynamic pruning#token-level inference#model efficiency
cs.CV2026

MixFrag: Fragility-Guided Mixed-Precision Post-Training Quantization for Vision Transformers

Md. Mehrab Hossain Opi, Robiul Islam Ryad, Md. Umar Faruk

MixFrag is a framework that estimates how sensitive each part of a Vision Transformer is to quantization and then assigns mixed-precision bit-widths using a knapsack optimization t…

#vision transformers#post-training quantization#mixed precision#model compression
cs.LG2026

Memory Efficient Tabular Foundation Models

Shuting Luo, Monika Mikhail Kanaan, Cameron Gordon +2

The paper studies how to reduce the memory footprint of tabular foundation models like TabPFN using compression techniques, achieving up to 7.6× memory savings with little performa…

#tabular data#foundation models#model compression#memory efficiency
cs.LG2026

Beyond the Best Teacher: Expanding and Compressing the Reasoning Solution Manifold

Songshuo Lu, Zhi Chen, Yaohua Tang

The paper proposes an expand‑then‑compress framework that builds a diverse set of RL‑trained teacher models and then distills them into a single student model, improving reasoning,…

#reinforcement learning#teacher‑student distillation#multi‑teacher aggregation#reasoning models
cs.LG2026

Prox: Training-Free FFN Activation Sparsity via Approximate Intermediate-Channel Salience in LLMs

Jinyi Liu, Wei Chen, Pengyu Chen +4

The paper introduces Prox, a training-free framework that sparsifies feed‑forward network activations in large language models by approximating intermediate‑channel salience, enabl…

#activation sparsity#feed-forward networks#large language models#training-free methods
cs.LG2026

OrthKD: Extracting Generalized Clinical Knowledge from Heterogeneous Teachers for Lightweight Deployment

Yi Xu, Cheng Chen, Mufan Cao

The paper introduces OrthKD, a knowledge distillation framework that selectively trusts a strong CNN and a weaker transformer teacher to train a lightweight MobileNetV3 model for d…

#knowledge distillation#diabetic retinopathy screening#edge deployment#heterogeneous teachers
cs.AI2026

How Small Can You Go? A Controlled Study of LoRA Rank, Target Modules, and Quantization Trade-offs for Text-to-SQL on a 60M-Parameter Model

Mahendra Singh Rathor, Anagheem Azzam

The paper investigates how LoRA rank, adapted modules, and low‑bit quantization trade off accuracy and resource usage when fine‑tuning a 60 M‑parameter T5‑small model for the WikiS…

#parameter-efficient fine-tuning#LoRA#quantization#text-to-sql
stat.ML2026

Gibbs randomness-compression proposition

M. Süzen, M Süzen

The paper proposes a theorem linking Gibbs entropy (a measure of randomness) to lossy model compression, showing that the entropy of remaining network weights correlates with learn…

#model compression#gibbs entropy#randomness#pruning
cs.LG2026

Zero-Shot Quantization for Object Detectors using Off-the-Shelf Generative Models

Hyunho Lee, Kyomin Hwang, Hyeonjin Kim +3

The paper proposes GoodQ, a method that uses off-the-shelf generative models to create synthetic training data for zero-shot quantization of object detectors, enabling low-bit quan…

#zero-shot quantization#object detection#generative models#model compression
cs.LG2026

Non-vacuous Generalization Bounds for Reinforcement Learning with Verifiable Rewards

Yuxuan Zhu, Rohan Alur, Daniel Kang

The paper derives the first non‑vacuous PAC‑Bayes generalization bounds for parameter‑efficient reinforcement learning with verifiable rewards applied to billion‑parameter language…

#reinforcement learning#generalization bounds#pac-bayes#large language models
cs.CL2026

Gold-Guided Programmatic Distillation for Financial Reasoning over Hybrid Tables and Text

Yun Dong, Erica Zhao, Elana Chen

The paper introduces a method that distills reliable numerical reasoning from a large language model into a smaller model by using execution-verified Python programs instead of tex…

#financial question answering#hybrid tables#programmatic distillation#numerical reasoning
cs.LG2026

DarwinLM: Evolutionary Structured Pruning of Large Language Models

Shengkun Tang, Oliver Sieberling, Eldar Kurtic +2

The paper introduces DarwinLM, an evolutionary search method for training-aware structured pruning of large language models that integrates lightweight post‑pruning training to fin…

#structured pruning#large language models#evolutionary algorithms#model compression
cs.LG2026

ExTernD: Expanded-Rank Ternary Decomposition Ternary LLM PTQ with Accuracy Approaching Any Quantization Level

Chethan Reddy G. P

The paper presents ExTernD, a post‑training factorization that expands the rank of ternary matrix decompositions to correct quantization errors, enabling large language models to a…

#quantization#large language models#ternary decomposition#post-training quantization
cs.LG2026

Constraint-Driven Model Optimization: An Industry Framework for Selecting Compression and Acceleration Techniques in Modern Machine Learning Systems

Dhruv Shivkant, Saket Mohanty, Utkarsh Wadhwa +1

The paper introduces a constraint-driven, multi-objective framework for selecting model compression and acceleration techniques based on deployment constraints such as latency, mem…

#model compression#inference optimization#constraint-driven selection#multi-objective engineering
cs.LG2026

Saturation Makes Quantization Error Additive: A Coverage Model with a Certificate

Joshua Hill

The paper demonstrates that for 4‑bit mixed‑precision quantization, the loss from quantizing a set of layers is largely additive across individual layers, and introduces a simple c…

#mixed-precision quantization#model compression#sensitivity analysis#coverage model
cs.CV2026

MobileSAM2: Lightweight Segment Anything for Spatial Intelligence

Kai Jiang, Jiaxing Huang, Jingyi Zhang +5

The paper introduces MobileSAM2, a lightweight version of the SAM2 segmentation model designed for mobile devices, using hypergraph-based knowledge distillation to transfer tempora…

#segment anything#mobile vision#knowledge distillation#hypergraph
cs.LG2026

dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats

Giuseppe Franco, Ian Colbert, Pablo Monteagudo-Lago +2

The paper presents dMX, a differentiable framework that learns per-layer floating‑point bit‑widths for large language models, enabling mixed‑precision quantization that balances ac…

#mixed-precision quantization#floating-point formats#large language models#model compression
cs.IR2026

Score-Only Distillation for Compact Dense Retrieval

Kirill Dubovikov, Martin Takac, Salem Lahlou

The paper proposes a method to compress large dense retrieval models into much smaller ones by distilling only the teacher's score vectors, using a row‑centered score‑vector object…

#dense retrieval#knowledge distillation#model compression#score-only training
cs.CV2026

Towards Efficient Convolutional Neural Network for Embedded Hardware via Multi-Dimensional Pruning

Hao Kong, Di Liu, Xiangzhong Luo +5

The paper introduces TECO, a framework that jointly prunes depth, width, and input resolution of convolutional neural networks to improve speed and resource usage on embedded devic…

#model compression#pruning#embedded hardware#convolutional neural networks
cs.LG2026

Prune, Update and Trim: Robust Structured Pruning for Large Language Models

Diego Coello de Portugal Mecke, Tom Hanika, Lars Schmidt-Thieme

The paper introduces Putri, a post‑training pruning method for large language models that updates remaining weights, prunes feed‑forward layers sequentially, and removes individual…

#structured pruning#large language models#post-training pruning#attention head pruning
cs.LG2026

Requential Coding: Pushing the Limits of Model Compression with Self-Generated Training Data

Shikai Qiu, Marc Finzi, Yujia Zheng +2

The paper proposes requential coding, a method where a teacher model selects training samples from the student’s own distribution so that only disagreements need to be encoded, yie…

#model compression#self-generated training data#PAC-Bayes bounds#large language models
cs.LG2026

Sparse Inter-Layer Dependencies of Transformer FFN Neurons

Johannes Knittel, Hanspeter Pfister

The paper introduces a training‑free method to attribute the activation of individual feed‑forward network neurons in Transformers to a small set of upstream neuron activations and…

#transformers#feedforward networks#neuron interpretability#sparsity
cs.CR2026

Securing LLMs in the Wild: Privacy and Security Challenges at the Edge

Ren-Yi Huang, Mingchen Li, Dumindu Samaraweera +1

The paper examines how efficiency‑focused optimizations for deploying large language models on edge devices create new security and privacy vulnerabilities, and it introduces a tax…

#edge computing#large language models#model compression#privacy attacks