# Reading is not Reasoning: Bridging the Agentic Policy Gap in Vision-Text Compression arXiv: https://arxiv.org/abs/2608.08960 ยท HTML: https://mlsanity.com/paper/2608.08960/ ## Key facts - Cited by 1 papers in the map (data through 14 Oct 2026). - Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. - Builds on 30 of its 42 references that are in the map. - First posted on arXiv 9 Aug 2026. ## Placed in - LLM agents and tool use / [Agent memory systems](https://mlsanity.com/agents/agent-memory-systems/) - Knowledge distillation / [Multimodal distillation](https://mlsanity.com/distill/multimodal-distillation/) ## Builds on (most cited first, 30 of 30) - [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024) - [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025) - [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022) - [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023) - [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023) - [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018) - [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023) - [TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension](https://mlsanity.com/paper/1705.03551/) (2017) - [HybridFlow: A Flexible and Efficient RLHF Framework](https://mlsanity.com/paper/2409.19256/) (2024) - [WebGPT: Browser-assisted question-answering with human feedback](https://mlsanity.com/paper/2112.09332/) (2021) - [Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning](https://mlsanity.com/paper/2503.09516/) (2025) - [Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps](https://mlsanity.com/paper/2011.01060/) (2020) - [ALFWorld: Aligning Text and Embodied Environments for Interactive Learning](https://mlsanity.com/paper/2010.03768/) (2020) - [When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories](https://mlsanity.com/paper/2212.10511/) (2022) - [DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence](https://mlsanity.com/paper/2606.19348/) (2026) - [MuSiQue: Multihop Questions via Single-hop Question Composition](https://mlsanity.com/paper/2108.00573/) (2021) - [Group-in-Group Policy Optimization for LLM Agent Training](https://mlsanity.com/paper/2505.10978/) (2025) - [Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe](https://mlsanity.com/paper/2604.13016/) (2026) - [MEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon Agents](https://mlsanity.com/paper/2506.15841/) (2025) - [MiMo-V2-Flash Technical Report](https://mlsanity.com/paper/2601.02780/) (2026) - [Reinforcement Learning Optimization for Large-Scale Learning: An Efficient and User-Friendly Scaling Library](https://mlsanity.com/paper/2506.06122/) (2025) - [SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization](https://mlsanity.com/paper/2604.02268/) (2026) - [Speculative Knowledge Distillation: Bridging the Teacher-Student Gap Through Interleaved Sampling](https://mlsanity.com/paper/2410.11325/) (2024) - [Glyph: Scaling Context Windows via Visual-Text Compression](https://mlsanity.com/paper/2510.17800/) (2025) - [AgentOCR: Reimagining Agent History via Optical Self-Compression](https://mlsanity.com/paper/2601.04786/) (2026) - [Leveraging Visual Tokens for Extended Text Contexts in Multi-Modal Learning](https://mlsanity.com/paper/2406.02547/) (2024) - [MemOCR: Layout-Aware Visual Memory for Efficient Long-Horizon Reasoning](https://mlsanity.com/paper/2601.21468/) (2026) - [LongSeeker: Elastic Context Orchestration for Long-Horizon Search Agents](https://mlsanity.com/paper/2605.05191/) (2026) - [See the Text: From Tokenization to Visual Reading](https://mlsanity.com/paper/2510.18840/) (2025) - [Reading, Not Thinking: Understanding and Bridging the Modality Gap When Text Becomes Pixels in Multimodal LLMs](https://mlsanity.com/paper/2603.09095/) (2026) ## Built on by (most cited first, 1 of 1) - [FocusVTC: Efficient and High-Performance Visual Text Compression with Adaptive Resolution](https://mlsanity.com/paper/2609.36651/) (2026) ## Abstract (arXiv) Multi-step language-model agents repeatedly process growing interaction histories, leading to substantial context costs. Vision--text compression reduces these costs by rendering history as images, but the resulting modality shift creates a marked capability gap. Through controlled evaluations of history recovery, matched-state decisions, and complete trajectories, we show that this gap cannot be explained by OCR quality alone. Visual-history agents exhibit systematic drift in action selection, query formulation, stopping, and evidence use, revealing an agentic policy gap. We introduce CAPS, a two-stage Cross-modal Agentic Policy Self-distillation framework that uses the same model's stronger text-history policy to supervise its visual-history counterpart. Offline trajectory self-distillation transfers successful text-policy behavior to visual-history inputs, while online policy self-distillation provides dense supervision on states visited by the visual-history policy during reinforcement learning. On SearchQA, CAPS improves over AgentOCR by 5.0 and 3.4 percentage points with 3B and 7B backbones, respectively. On full-history ALFWorld, the corresponding gains are 15.6 and 14.5 percentage points. Across settings, CAPS reduces average memory-context cost by up to 63.3% and peak cost by up to 83.4% relative to matched text-history policies. These results show that explicit cross-modal policy self-distillation can preserve agent capability under vision--text compression. Our code will be made publicly available.