Communication between Frozen Large Language Models via Prompt Optimization in a Referential Game

arXiv:2609.31989 · cs.CL, cs.AI, cs.LG, cs.MA · Submitted 2026-09-25 · Read on arXiv

cs.CL, cs.AI, cs.LG, cs.MA

Submitted: 2026-09-25

Updated: 2026-09-25

Code: https://github.com/a2aproject/a2a

Terminology

• Accuracy — 27× in this paper · explained in 1 episode
• Decode — 14× in this paper · explained in 2 episodes
• Frozen model — 8× in this paper · explained in 3 episodes
• Generalization — 4× in this paper · explained in 13 episodes
• Ablation — 2× in this paper · explained in 1 episode
• Consistency — 2× in this paper · explained in 3 episodes
• Hamming Distance — 2× in this paper · explained in 2 episodes
• Tokenization — 2× in this paper · explained in 2 episodes
• Trace — 2× in this paper · explained in 2 episodes
• Agentic AI — 1× in this paper · explained in 1 episode
• Agentic System — 1× in this paper · explained in 2 episodes
• Calibration — 1× in this paper · explained in 7 episodes
• Cooperation — 1× in this paper · explained in 2 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes
• Provenance — 1× in this paper · explained in 4 episodes
• Quantization — 1× in this paper · explained in 12 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Robustness — 1× in this paper · explained in 5 episodes

Sources

• A Scalable Communication Protocol for Networks of Large Language Models
• Beyond Natural Language: LLMs Leveraging Alternative Formats for Enhanced Reasoning and Communication
• Multi-Agent Cooperation and the Emergence of (Natural) Language
• Compositionality and Generalization in Emergent Languages
• Searching for Structure: Investigating Emergent Communication with Large Language Models
• From Signals to Structure: How Memory Architecture Drives Language Emergence in LLM Agents
• Emergent Languages in Populations of Language Model Agents: From Token Efficiency to Oversight Evasion
• Compositional Languages Emerge in a Neural Iterated Learning Model
• Large Language Models as Optimizers
• Automatic Prompt Optimization with "Gradient Descent" and Beam Search
• TextGrad: Automatic "Differentiation" via Text
• GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning
• Automated Design of Agentic Systems
• Optimas: Optimizing Compound AI Systems with Globally Aligned Local Rewards
• GLM-5: from Vibe Coding to Agentic Engineering
• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering