NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness

arXiv:2609.08183 · cs.CL · Submitted 2026-09-08 · Read on arXiv

cs.CL

Submitted: 2026-09-08

Updated: 2026-10-14

Comments: Huggingface: https://hf.co/collections/TokenRhythm/neohorse-1; Github: https://github.com/TokenRhythm/NeoHorse

Code: https://github.com/TokenRhythm/NeoHorse

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Benchmark — 1× in this paper · explained in 2 episodes
• Distillation — 1× in this paper · explained in 2 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• On-Policy Distillation — 1× in this paper · explained in 3 episodes

Sources

• Scaling the Horizon, Not the Parameters: Reaching Trillion-Parameter Performance with a 35B Agent
• $\tau^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment
• FireAct: Toward Language Agent Fine-tuning
• FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance
• Evaluating Large Language Models Trained on Code
• Recursive Self-Improvement in AI: From Bounded Self-Refinement to Autonomous Research Loops
• Agent-FLAN: Designing Data and Methods of Effective Agent Tuning for Large Language Models
• Co-Harness: Co-Evolving Harnesses and Model Weights for LLM Agents
• DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
• ReTool: Reinforcement Learning for Strategic Tool Use in LLMs
• The Llama 3 Herd of Models
• VitaBench: Benchmarking LLM Agents with Versatile Interactive Tasks in Real-world Applications
• Distilling the Knowledge in a Neural Network
• LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code
• Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning
• Continual Harness: Online Adaptation for Self-Improving Foundation Agents
• The Interplay of Harness Design and Post-Training in LLM Agents
• Nanbeige4.2-3B: Unlocking Agentic Capabilities in a Compact Model

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