# LLMs can see and hear without any training

arXiv: https://arxiv.org/abs/2501.18096 · HTML: https://mlsanity.com/paper/2501.18096/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 30 Jan 2025.

## Built on by (most cited first, 8 of 8)

- [Learning to See Before Seeing: Demystifying LLM Visual Priors from Language Pre-training](https://mlsanity.com/paper/2509.26625/) (2025)
- [It's Never Too Late: Noise Optimization for Collapse Recovery in Trained Diffusion Models](https://mlsanity.com/paper/2601.00090/) (2025)
- [Words That Make Language Models Perceive](https://mlsanity.com/paper/2510.02425/) (2025)
- [Exploiting Mixture-of-Experts Redundancy Unlocks Multimodal Generative Abilities](https://mlsanity.com/paper/2503.22517/) (2025)
- [DUALVISION: RGB-Infrared Multimodal Large Language Models for Robust Visual Reasoning](https://mlsanity.com/paper/2604.18829/) (2026)
- [No Concept Left Behind: Test-Time Optimization for Compositional Text-to-Image Generation](https://mlsanity.com/paper/2509.23457/) (2025)
- [Personalizing Text-to-Image Generation to Individual Taste](https://mlsanity.com/paper/2604.07427/) (2026)
- [LINE: LLM-based Iterative Neuron Explanations for Vision Models](https://mlsanity.com/paper/2604.08039/) (2026)

## Abstract (arXiv)

We present MILS: Multimodal Iterative LLM Solver, a surprisingly simple, training-free approach, to imbue multimodal capabilities into your favorite LLM. Leveraging their innate ability to perform multi-step reasoning, MILS prompts the LLM to generate candidate outputs, each of which are scored and fed back iteratively, eventually generating a solution to the task. This enables various applications that typically require training specialized models on task-specific data. In particular, we establish a new state-of-the-art on emergent zero-shot image, video and audio captioning. MILS seamlessly applies to media generation as well, discovering prompt rewrites to improve text-to-image generation, and even edit prompts for style transfer! Finally, being a gradient-free optimization approach, MILS can invert multimodal embeddings into text, enabling applications like cross-modal arithmetic.