arXiv AI
AgentOCR: Reimagining Agent History via Optical Self-Compression
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv Computation and Language
Sep 1
When History Is Multimodal: Rethinking Context Management for Long-Horizon Agents
arXiv:2608.29897v1 Announce Type: new Abstract: Long-horizon agents need a context manager to compress growing interaction histories into a bounded working context, via passive strategies or active s...
arXiv AI
Aug 11
Reading is not Reasoning: Bridging the Agentic Policy Gap in Vision-Text Compression
arXiv:2608. 08960v1 Announce Type: new Abstract: Multi-step language-model agents repeatedly process growing interaction histories, leading to substantial context costs.
arXiv Computation and Language
3d ago
ReVision: Scaling Computer-Use Agents via Temporal Visual Redundancy Reduction
arXiv:2605.11212v4 Announce Type: replace Abstract: Computer-use agents (CUAs) rely on visual observations of graphical user interfaces, where each screenshot is encoded into a large number of visual...
arXiv Computation and Language
Oct 9
GUI-KV: Efficient GUI Agents via KV Cache with Spatio-Temporal Awareness
GUI-KV is a plug‑and‑play key‑value cache compression method designed for graphical user interface agents that use vision‑language models. It exploits spatial saliency and temporal redundancy to prune cache entries, achieving near‑full‑cache accuracy while using modest memory budgets. Experiments on standard GUI agent benchmarks show significant reductions in decoding FLOPs and improvements in step accuracy compared to existing compression baselines.
arXiv Computer Vision
Sep 11
Caption-once, Frames-on-Demand: Visual-Need Routing for Budget-Aware Agentic Long Video Understanding
The paper introduces Caption‑once, Frames‑on‑Demand (CFD), a budget‑aware edge‑cloud framework for long‑video understanding. CFD first runs a single offline captioning pass on the edge to build a dual‑track narrative index—an event‑level story skeleton and a clip‑level micro‑log—that is cached for future queries. At query time, a cloud‑side MLLM uses a Visual‑Need Router to decide whether to retrieve keyframes for perceptual questions, thereby limiting visual processing while preserving temporal structure in language space.
arXiv Computer Vision
Oct 8
VideoEvolve: Co-Evolving Memory and Retrieval for Long Video Understanding
VideoEvolve is a self‑evolving framework for long‑video understanding that jointly updates both memory and retrieval mechanisms. It starts with a coarse low‑frame‑rate overview and uses a Memory Evolver to selectively augment memory while a Retrieval Evolver adapts how information is fetched from that memory. The two components are alternately refined via Agentic Reinforcement Learning, guided by Bottleneck‑Aware and Capability‑Aware Evolution Feedback to focus on the most limiting aspect and to prevent over‑specialization, respectively. Experiments on several long‑video benchmarks show that this approach improves performance over static systems.