# VisualClaw: A Real-Time, Personalized Agent for the Physical World

arXiv: https://arxiv.org/abs/2606.16295 · HTML: https://mlsanity.com/paper/2606.16295/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Embodied and 3D vision-language models, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 30 of its 44 references that are in the map.
- First posted on arXiv 15 Jun 2026.

## Placed in

- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)
- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

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

- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [Voyager: An Open-Ended Embodied Agent with Large Language Models](https://mlsanity.com/paper/2305.16291/) (2023)
- [Ego4D: Around the World in 3,000 Hours of Egocentric Video](https://mlsanity.com/paper/2110.07058/) (2021)
- [Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video Understanding](https://mlsanity.com/paper/2306.02858/) (2023)
- [ExpeL: LLM Agents Are Experiential Learners](https://mlsanity.com/paper/2308.10144/) (2023)
- [NExT-QA:Next Phase of Question-Answering to Explaining Temporal Actions](https://mlsanity.com/paper/2105.08276/) (2021)
- [EgoSchema: A Diagnostic Benchmark for Very Long-form Video Language Understanding](https://mlsanity.com/paper/2308.09126/) (2023)
- [Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces](https://mlsanity.com/paper/2412.14171/) (2024)
- [MovieChat: From Dense Token to Sparse Memory for Long Video Understanding](https://mlsanity.com/paper/2307.16449/) (2023)
- [RL2: Fast Reinforcement Learning via Slow Reinforcement Learning](https://mlsanity.com/paper/1611.02779/) (2016)
- [TimeChat: A Time-sensitive Multimodal Large Language Model for Long Video Understanding](https://mlsanity.com/paper/2312.02051/) (2023)
- [Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables](https://mlsanity.com/paper/1903.08254/) (2019)
- [VideoAgent: Long-form Video Understanding with Large Language Model as Agent](https://mlsanity.com/paper/2403.10517/) (2024)
- [Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense Video Captioning](https://mlsanity.com/paper/2302.14115/) (2023)
- [VTimeLLM: Empower LLM to Grasp Video Moments](https://mlsanity.com/paper/2311.18445/) (2023)
- [VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos](https://mlsanity.com/paper/2405.19209/) (2024)
- [MA-LMM: Memory-Augmented Large Multimodal Model for Long-Term Video Understanding](https://mlsanity.com/paper/2404.05726/) (2024)
- [Self-Chained Image-Language Model for Video Localization and Question Answering](https://mlsanity.com/paper/2305.06988/) (2023)
- [A Simple LLM Framework for Long-Range Video Question-Answering](https://mlsanity.com/paper/2312.17235/) (2023)
- [Actor and Observer: Joint Modeling of First and Third-Person Videos](https://mlsanity.com/paper/1804.09627/) (2018)
- [ProMP: Proximal Meta-Policy Search](https://mlsanity.com/paper/1810.06784/) (2018)
- [Frame-Voyager: Learning to Query Frames for Video Large Language Models](https://mlsanity.com/paper/2410.03226/) (2024)
- [EgoThink: Evaluating First-Person Perspective Thinking Capability of Vision-Language Models](https://mlsanity.com/paper/2311.15596/) (2023)
- [MetaClaw: Just Talk -- An Agent That Meta-Learns and Evolves in the Wild](https://mlsanity.com/paper/2603.17187/) (2026)
- [A Simple Recipe for Contrastively Pre-training Video-First Encoders Beyond 16 Frames](https://mlsanity.com/paper/2312.07395/) (2023)
- [VidEgoThink: Assessing Egocentric Video Understanding Capabilities for Embodied AI](https://mlsanity.com/paper/2410.11623/) (2024)
- [VisionClaw: Always-On AI Agents through Smart Glasses](https://mlsanity.com/paper/2604.03486/) (2026)

## Built on by (most cited first, 3 of 3)

- [LightMem-Ego: Your AI Memory for Everyday Life](https://mlsanity.com/paper/2607.11487/) (2026)
- [Chain-of-Experience for Continual LLM Improvement](https://mlsanity.com/paper/2608.18027/) (2026)
- [VideoEvoHarness: Self Evolving Agent Harness from Unlabeled Videos for Video Understanding](https://mlsanity.com/paper/2610.15389/) (2026)

## Abstract (arXiv)

Vision language models are serving as general-purpose interfaces for complex multimodal tasks. However, deployment still faces three gaps: VLMs typically incur high latency and cost when processing dense video frames and long prompts, the agent scaffold remains static after deployment, and standard video-QA benchmarks do not test whether agents can use visual evidence inside tool-using workspaces. We present VisualClaw, a self-evolving multimodal agent built around two principles. First, hybrid encoding reduces deployment cost by filtering less informative streaming frames with a cascaded gate and compressing the text skill bank through hot/cold top-k injection. Second, skill evolution lets the agent learn from failures: retrieved memories condition an evolver as direct concatenated context or as guided evidence, producing skill-bank updates that help future questions. Across 4 video-QA benchmarks with 2 VLMs, VisualClaw cuts per-question API cost by an average -98% versus full-frame upload and by -25.9% over the offline uniform 8 frame baseline, while boosting accuracy in most settings, e.g., an average +3.85% and a peak +15.80% on EgoSchema with Gemini 3 Flash. To address the gap, we curate VisualClawArena, a 200-scenario multimodal agentic benchmark built through a strict five-stage pipeline; models must use video evidence, documents, dynamic updates, and executable checks inside a workspace. On VisualClawArena, the same framework with computer-use agent backends improves macro accuracy by +2.9% for Codex (GPT-5.5) and +3.2% for Claude Code (Sonnet 4.6) over no-evolution baselines, with a -9.5% cost reduction compared to the uniform-sampled baseline. These properties make VisualClaw a natural fit for edge applications, where the cascade reduces a 1-hour streaming session from ~3,600 API uploads down to only 5-20 calls and the self-evolution makes it a perfect personalized assistant.