# Making VLMs More Robot-Friendly: Self-Critical Distillation of Low-Level Procedural Reasoning

arXiv: https://arxiv.org/abs/2507.08224 · HTML: https://mlsanity.com/paper/2507.08224/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning.
- Builds on 19 of its 51 references that are in the map.
- First posted on arXiv 11 Jul 2025.

## Placed in

- LLM reasoning / [Spatial and embodied reasoning](https://mlsanity.com/lr/spatial-and-embodied-reasoning/)

## Builds on (most cited first, 19 of 19)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models](https://mlsanity.com/paper/2304.10592/) (2023)
- [PaLM-E: An Embodied Multimodal Language Model](https://mlsanity.com/paper/2303.03378/) (2023)
- [Inner Monologue: Embodied Reasoning through Planning with Language Models](https://mlsanity.com/paper/2207.05608/) (2022)
- [Gemma: Open Models Based on Gemini Research and Technology](https://mlsanity.com/paper/2403.08295/) (2024)
- [PaliGemma: A versatile 3B VLM for transfer](https://mlsanity.com/paper/2407.07726/) (2024)
- [CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation](https://mlsanity.com/paper/2411.19650/) (2024)
- [NaVILA: Legged Robot Vision-Language-Action Model for Navigation](https://mlsanity.com/paper/2412.04453/) (2024)
- [Vision-Language Models Can Self-Improve Reasoning via Reflection](https://mlsanity.com/paper/2411.00855/) (2024)
- [Enhancing LLM Reasoning via Critique Models with Test-Time and Training-Time Supervision](https://mlsanity.com/paper/2411.16579/) (2024)
- [Guiding Long-Horizon Task and Motion Planning with Vision Language Models](https://mlsanity.com/paper/2410.02193/) (2024)
- [RLCD: Reinforcement Learning from Contrastive Distillation for Language Model Alignment](https://mlsanity.com/paper/2307.12950/) (2023)
- [VILA2: VILA Augmented VILA](https://mlsanity.com/paper/2407.17453/) (2024)
- [Automating Robot Failure Recovery Using Vision-Language Models With Optimized Prompts](https://mlsanity.com/paper/2409.03966/) (2024)
- [MFE-ETP: A Comprehensive Evaluation Benchmark for Multi-modal Foundation Models on Embodied Task Planning](https://mlsanity.com/paper/2407.05047/) (2024)
- [Survey on Vision-Language-Action Models](https://mlsanity.com/paper/2502.06851/) (2025)
- [SDRT: Enhance Vision-Language Models by Self-Distillation with Diverse Reasoning Traces](https://mlsanity.com/paper/2503.01754/) (2025)

## Built on by (most cited first, 4 of 4)

- [EmbodiedBrain: Expanding Performance Boundaries of Task Planning for Embodied Intelligence](https://mlsanity.com/paper/2510.20578/) (2025)
- [RoboEvolve: Co-Evolving Planner-Simulator for Robotic Manipulation with Limited Data](https://mlsanity.com/paper/2605.13775/) (2026)
- [BodyCam-VQA: Enhanced Body-Worn Camera Video Captioning via Multimodal Reasoning and Probe Question Generation](https://mlsanity.com/paper/2609.10815/) (2026)
- [EmbodiedMind: Adaptive Data Curation and Prefix-Tree Reinforcement Learning for Efficient Embodied Intelligence](https://mlsanity.com/paper/2609.19659/) (2026)

## Abstract (arXiv)

Large language models (LLMs) have shown promise in robotic procedural planning, yet their human-centric reasoning often omits the low-level, grounded details needed for robotic execution. Vision-language models (VLMs) offer a path toward more perceptually grounded plans, but current methods either rely on expensive, large-scale models or are constrained to narrow simulation settings. We introduce SelfReVision, a lightweight and scalable self-improvement framework for vision-language procedural planning. SelfReVision enables small VLMs to iteratively critique, revise, and verify their own plans-without external supervision or teacher models-drawing inspiration from chain-of-thought prompting and self-instruct paradigms. Through this self-distillation loop, models generate higher-quality, execution-ready plans that can be used both at inference and for continued fine-tuning. Using models varying from 3B to 72B, our results show that SelfReVision not only boosts performance over weak base VLMs but also outperforms models 100X the size, yielding improved control in downstream embodied tasks.