# Homer: Understanding Long-form Videos with Hierarchical Memory and Agentic Reasoning

arXiv: https://arxiv.org/abs/2607.02588 · HTML: https://mlsanity.com/paper/2607.02588/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 13 of its 37 references that are in the map.
- First posted on arXiv 1 Jul 2026.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [Video understanding with MLLMs](https://mlsanity.com/mllm/video-understanding-with-mllms/)
- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 13 of 13)

- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [mimic-video: Video-Action Models for Generalizable Robot Control Beyond VLAs](https://mlsanity.com/paper/2512.15692/) (2025)
- [LongVT: Incentivizing "Thinking with Long Videos" via Native Tool Calling](https://mlsanity.com/paper/2511.20785/) (2025)
- [StreamMem: Query-Agnostic KV Cache Memory for Streaming Video Understanding](https://mlsanity.com/paper/2508.15717/) (2025)
- [StreamingTOM: Streaming Token Compression for Efficient Video Understanding](https://mlsanity.com/paper/2510.18269/) (2025)
- [FOCUS: Efficient Keyframe Selection for Long Video Understanding](https://mlsanity.com/paper/2510.27280/) (2025)
- [From Seconds to Hours: Reviewing MultiModal Large Language Models on Comprehensive Long Video Understanding](https://mlsanity.com/paper/2409.18938/) (2024)
- [FluxMem: Adaptive Hierarchical Memory for Streaming Video Understanding](https://mlsanity.com/paper/2603.02096/) (2026)
- [From Verbatim to Gist: Distilling Pyramidal Multimodal Memory via Semantic Information Bottleneck for Long-Horizon Video Agents](https://mlsanity.com/paper/2603.01455/) (2026)
- [CL-bench Life: Can Language Models Learn from Real-Life Context?](https://mlsanity.com/paper/2604.27043/) (2026)
- [Revisiting Long-context Modeling from Context Denoising Perspective](https://mlsanity.com/paper/2510.05862/) (2025)
- [FocusGraph: Graph-Structured Frame Selection for Embodied Long Video Question Answering](https://mlsanity.com/paper/2603.04349/) (2026)
- [PyraVid: Hierarchical Multimodal Memory for Long-Horizon Video Reasoning](https://mlsanity.com/paper/2605.17065/) (2026)

## Built on by (most cited first, 2 of 2)

- [VideoHarness-RSI: Recursive Harness Self-Improvement for Long-Video Understanding with Frozen Vision-Language Models](https://mlsanity.com/paper/2608.24302/) (2026)
- [VideoEvoHarness: Self Evolving Agent Harness from Unlabeled Videos for Video Understanding](https://mlsanity.com/paper/2610.15389/) (2026)

## Abstract (arXiv)

Multimodal large language models excel on short clips but struggle on hour-long videos in an online setting, where frames are processed incrementally under limited memory. Existing online methods either retain compact visual representations that lack semantic structure, or build higher-level memory stores organized around temporal proximity rather than explicit causal links, leaving multi-hop narrative reasoning to be reconstructed by the LLM at every query. We bridge this gap with textscHomer, a Hierarchical Online Memory Exploration and Reasoning framework. textscHomer's memory mirrors the multi-scale structure of long videos, ranging from raw perception, to recurring entities, to events connected by explicit temporal and causal relations. Its agentic reasoner then explores this memory the way humans do, locating the relevant scene, looking up details, and composing the answer through multi-round memory retrieval, with a harness that verifies and corrects each step. textscHomer outperforms the previous best agent method by +5.5, +10.8, and +4.4 points on M3-Bench-robot, M3-Bench-web, and Video-MME-Long, and consistently lifts three various LLM backbones, indicating a model-agnostic structural capability for grounded retrieval over long videos.