Seeing What Matters, Verifying What Pays: Load-Adaptive Speculative Decoding for Vision-Language Models
cs.AI, cs.CV
Submitted: 2026-11-01
Updated: 2026-11-06
Code: https://github.com/sahil280114/codealpaca
Terminology
Sources
- Fast Transformer Decoding: One Write-Head is All You Need
- Accelerating Large Language Model Decoding with Speculative Sampling
- Speculative Decoding Reimagined for Multimodal Large Language Models
- Is Multimodal Speculative Decoding Ready for Diffusion-Based Parallel Drafting? A Survey and Empirical Diagnosis
- SpecVLM: Fast Speculative Decoding in Vision-Language Models
- Vision Is Not Overhead: One-Pass Block Drafting for Lossless Speculative Decoding in Vision-Language Models
- Sequoia: Scalable, Robust, and Hardware-aware Speculative Decoding
- Accelerating Speculative Decoding with Block Diffusion Draft Trees
- Cost-Aware Diffusion Draft Trees for Speculative Decoding
- FlexDraft: Flexible Speculative Decoding via Attention Tuning and Bonus-Guided Calibration
- MMSpec: Benchmarking Speculative Decoding for Vision-Language Models
- Recurrent Drafter for Fast Speculative Decoding in Large Language Models
- Correctness Forensics for Batch Speculative Decoding: Diagnosing the Ragged Tensor Problem
- SPECTRE: Hybrid Ordinary-Parallel Speculative Serving for Resource-Efficient LLM Inference
- Training Verifiers to Solve Math Word Problems
- Evaluating Large Language Models Trained on Code
- Program Synthesis with Large Language Models
- olmOCR: Unlocking Trillions of Tokens in PDFs with Vision Language Models
- Microsoft COCO Captions: Data Collection and Evaluation Server
- Pixtral 12B
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents