# RECIPE: Procedural Planning via Grounding in Instructional Video

arXiv: https://arxiv.org/abs/2605.19976 · HTML: https://mlsanity.com/paper/2605.19976/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 42 of its 50 references that are in the map.
- First posted on arXiv 19 May 2026.

## Builds on (most cited first, 30 of 42)

- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [OpenVLA: An Open-Source Vision-Language-Action Model](https://mlsanity.com/paper/2406.09246/) (2024)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [Ego4D: Around the World in 3,000 Hours of Egocentric Video](https://mlsanity.com/paper/2110.07058/) (2021)
- [Tulu 3: Pushing Frontiers in Open Language Model Post-Training](https://mlsanity.com/paper/2411.15124/) (2024)
- [Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents](https://mlsanity.com/paper/2201.07207/) (2022)
- [HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips](https://mlsanity.com/paper/1906.03327/) (2019)
- [End-to-End Learning of Visual Representations from Uncurated Instructional Videos](https://mlsanity.com/paper/1912.06430/) (2019)
- [Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives](https://mlsanity.com/paper/2311.18259/) (2023)
- [Video-R1: Reinforcing Video Reasoning in MLLMs](https://mlsanity.com/paper/2503.21776/) (2025)
- [Understanding the Effects of RLHF on LLM Generalisation and Diversity](https://mlsanity.com/paper/2310.06452/) (2023)
- [COIN: A Large-scale Dataset for Comprehensive Instructional Video Analysis](https://mlsanity.com/paper/1903.02874/) (2019)
- [Project Aria: A New Tool for Egocentric Multi-Modal AI Research](https://mlsanity.com/paper/2308.13561/) (2023)
- [Defining and Characterizing Reward Hacking](https://mlsanity.com/paper/2209.13085/) (2022)
- [HoloAssist: an Egocentric Human Interaction Dataset for Interactive AI Assistants in the Real World](https://mlsanity.com/paper/2309.17024/) (2023)
- [Unsupervised Learning from Narrated Instruction Videos](https://mlsanity.com/paper/1506.09215/) (2015)
- [Procedure Planning in Instructional Videos](https://mlsanity.com/paper/1907.01172/) (2019)
- [My View is the Best View: Procedure Learning from Egocentric Videos](https://mlsanity.com/paper/2207.10883/) (2022)
- [EgoPlan-Bench: Benchmarking Multimodal Large Language Models for Human-Level Planning](https://mlsanity.com/paper/2312.06722/) (2023)
- [P3IV: Probabilistic Procedure Planning from Instructional Videos with Weak Supervision](https://mlsanity.com/paper/2205.02300/) (2022)
- [CaptainCook4D: A Dataset for Understanding Errors in Procedural Activities](https://mlsanity.com/paper/2312.14556/) (2023)
- [Learning Procedure-aware Video Representation from Instructional Videos and Their Narrations](https://mlsanity.com/paper/2303.17839/) (2023)
- [EgoPlan-Bench2: A Benchmark for Multimodal Large Language Model Planning in Real-World Scenarios](https://mlsanity.com/paper/2412.04447/) (2024)
- [Drop-DTW: Aligning Common Signal Between Sequences While Dropping Outliers](https://mlsanity.com/paper/2108.11996/) (2021)
- [StepFormer: Self-supervised Step Discovery and Localization in Instructional Videos](https://mlsanity.com/paper/2304.13265/) (2023)
- [PDPP: Projected Diffusion for Procedure Planning in Instructional Videos](https://mlsanity.com/paper/2303.14676/) (2023)
- [Planning with Reasoning using Vision Language World Model](https://mlsanity.com/paper/2509.02722/) (2025)
- [HowToCaption: Prompting LLMs to Transform Video Annotations at Scale](https://mlsanity.com/paper/2310.04900/) (2023)

## Abstract (arXiv)

Visual planning asks a model to generate the remaining steps of a procedure in natural language given a partial video context and a goal. Progress on this task is bottlenecked by annotation: clean labeled datasets are small, domain-narrow, and encode a single execution trajectory per example, even though many valid orderings exist. Large-scale instructional video corpora offer orders of magnitude more procedural content, but supervised fine-tuning on pseudo-labels from their noisy ASR narrations propagates segmentation and alignment errors and stays single-trajectory. We identify a key asymmetry: extracting clean step labels from noisy video is hard, but verifying whether a generated step sequence is temporally grounded in ASR transcripts is cheap and scales to millions of videos via precomputed text embeddings. We exploit this asymmetry in RECIPE, which uses grounding quality as a reward for GRPO, turning the noisy corpus into a verifier rather than a label source. The framework applies uniformly to two planner input configurations (Socratic, with a textual history extracted by a frozen VLM, and Video, consuming video tokens directly) and to annotated and weakly supervised regimes. We evaluate on 7 procedural benchmarks using a reference-based LLM-as-judge protocol scoring plans across 6 procedural criteria. RECIPE-RL improves over the base checkpoint at all scales (0.5B, 3B, 7B) and every benchmark, with macro-accuracy gains of +7 to +8 points in-domain and up to +16 points zero-shot. It outperforms supervised fine-tuning on both annotated and pseudo-labeled plans (the latter degrades the base) and remains robust without human annotations. Used as the proposal stage of a prior propose-assess-search planner, it improves over the strongest zero-shot baseline at every horizon on Visual Planning for Assistance, and on COIN it preserves the generation diversity that SFT collapses.