Octopus: One Brain, Many Branches for Latent-KV Multi-Agent Reinforcement Learning
cs.LG, cs.CL
Submitted: 2026-11-03
Updated: 2026-11-06
Code: https://github.com/shatianming5/octopus
Terminology
Sources
- OpenResearcher: A Fully Open Pipeline for Long-Horizon Deep Research Trajectory Synthesis
- Cache Engineering for Agents
- ReSum: Unlocking Long-Horizon Search Intelligence via Context Summarization
- ACM: Agentic Context Management for Long Horizon Tasks
- Context Language Models
- KVCMAS: Efficient KV cache Correction for Shared Context in Multi-Agent Systems
- Learning Adaptive Parallel Reasoning with Language Models
- Multi-Agent Deep Research: Training Multi-Agent Systems with M-GRPO
- What Attention Recalls and Recurrence Controls in Hybrid Language Models
- ActiveMem: Dynamic Latent Memory Trees for Long-Horizon Agents
- Self-Compacting Language Model Agents
- Recursive Language Models
- MEMENTO: Teaching LLMs to Manage Their Own Context
- Learning to Communicate: Toward End-to-End Optimization of Multi-Agent Language Systems
- AREAL-DTA: Dynamic Tree Attention for Efficient Reinforcement Learning of Large Language Models
- QwenGyre: An Elastic Reinforcement Learning Framework for Training xLong-Horizon Agents
- MultiAgent-OPD: End-to-End Training of Hierarchical Multi-Agent Systems with On-Policy Distillation
- Understanding R1-Zero-Like Training: A Critical Perspective
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis