# Atom: Efficient On-Device Video-Language Pipelines Through Modular Reuse

arXiv: https://arxiv.org/abs/2512.17108 · HTML: https://mlsanity.com/paper/2512.17108/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 37 references that are in the map.
- First posted on arXiv 18 Dec 2025.

## Builds on (most cited first, 22 of 22)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration](https://mlsanity.com/paper/2306.00978/) (2023)
- [Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval](https://mlsanity.com/paper/2104.00650/) (2021)
- [CLIP4Clip: An Empirical Study of CLIP for End to End Video Clip Retrieval](https://mlsanity.com/paper/2104.08860/) (2021)
- [NExT-GPT: Any-to-Any Multimodal LLM](https://mlsanity.com/paper/2309.05519/) (2023)
- [Panda-70M: Captioning 70M Videos with Multiple Cross-Modality Teachers](https://mlsanity.com/paper/2402.19479/) (2024)
- [InternVideo2: Scaling Foundation Models for Multimodal Video Understanding](https://mlsanity.com/paper/2403.15377/) (2024)
- [The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits](https://mlsanity.com/paper/2402.17764/) (2024)
- [mPLUG-2: A Modularized Multi-modal Foundation Model Across Text, Image and Video](https://mlsanity.com/paper/2302.00402/) (2023)
- [Localizing Moments in Video with Temporal Language](https://mlsanity.com/paper/1809.01337/) (2018)
- [LLM Pruning and Distillation in Practice: The Minitron Approach](https://mlsanity.com/paper/2408.11796/) (2024)
- [Spotlight: Mobile UI Understanding using Vision-Language Models with a Focus](https://mlsanity.com/paper/2209.14927/) (2022)
- [Fast On-device LLM Inference with NPUs](https://mlsanity.com/paper/2407.05858/) (2024)
- [VideoRAG: Retrieval-Augmented Generation over Video Corpus](https://mlsanity.com/paper/2501.05874/) (2025)
- [Any-Precision LLM: Low-Cost Deployment of Multiple, Different-Sized LLMs](https://mlsanity.com/paper/2402.10517/) (2024)
- [LLMCad: Fast and Scalable On-device Large Language Model Inference](https://mlsanity.com/paper/2309.04255/) (2023)
- [Retrieval-Augmented Visual Question Answering via Built-in Autoregressive Search Engines](https://mlsanity.com/paper/2502.16641/) (2025)
- [TG-LLaVA: Text Guided LLaVA via Learnable Latent Embeddings](https://mlsanity.com/paper/2409.09564/) (2024)
- [VVS: Video-to-Video Retrieval with Irrelevant Frame Suppression](https://mlsanity.com/paper/2303.08906/) (2023)
- [Towards Retrieval Augmented Generation over Large Video Libraries](https://mlsanity.com/paper/2406.14938/) (2024)
- [Text-Video Multi-Grained Integration for Video Moment Montage](https://mlsanity.com/paper/2412.09276/) (2024)

## Abstract (arXiv)

Recent advances in video-language models have enabled powerful applications like video retrieval, captioning, and assembly. However, executing such multi-stage pipelines efficiently on mobile devices remains challenging due to redundant model loads and fragmented execution. We introduce Atom, an on-device system that restructures video-language pipelines for fast and efficient execution. Atom decomposes a billion-parameter model into reusable modules, such as the visual encoder and language decoder, and reuses them across subtasks like captioning, reasoning, and indexing. This reuse-centric design eliminates repeated model loading and enables parallel execution, reducing end-to-end latency without sacrificing performance. On commodity smartphones, Atom achieves 27--33% faster execution compared to non-reuse baselines, with only marginal performance drop (≤ 2.3 Recall@1 in retrieval, ≤ 1.5 CIDEr in captioning). These results position Atom as a practical, scalable approach for efficient video-language understanding on edge devices.