# Training Agents to Evolve with Their Harness: TaoLive Digital Avatar Agent Technical Report

arXiv: https://arxiv.org/abs/2608.15763 · HTML: https://mlsanity.com/paper/2608.15763/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use.
- Builds on 25 of its 43 references that are in the map.
- First posted on arXiv 16 Aug 2026.

## Placed in

- LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/)

## Builds on (most cited first, 25 of 25)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://mlsanity.com/paper/2503.14476/) (2025)
- [Tulu 3: Pushing Frontiers in Open Language Model Post-Training](https://mlsanity.com/paper/2411.15124/) (2024)
- [DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence](https://mlsanity.com/paper/2606.19348/) (2026)
- [On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes](https://mlsanity.com/paper/2306.13649/) (2023)
- [A Comprehensive Survey of Continual Learning: Theory, Method and Application](https://mlsanity.com/paper/2302.00487/) (2023)
- [Group Sequence Policy Optimization](https://mlsanity.com/paper/2507.18071/) (2025)
- [DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines](https://mlsanity.com/paper/2310.03714/) (2023)
- [EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty](https://mlsanity.com/paper/2401.15077/) (2024)
- [A Comprehensive Survey on Test-Time Adaptation under Distribution Shifts](https://mlsanity.com/paper/2303.15361/) (2023)
- [GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization](https://mlsanity.com/paper/2601.05242/) (2026)
- [Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses](https://mlsanity.com/paper/2604.25850/) (2026)
- [DFlash: Block Diffusion for Flash Speculative Decoding](https://mlsanity.com/paper/2602.06036/) (2026)
- [Self-Harness: Harnesses That Improve Themselves](https://mlsanity.com/paper/2606.09498/) (2026)
- [Code as Agent Harness](https://mlsanity.com/paper/2605.18747/) (2026)
- [DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation](https://mlsanity.com/paper/2607.05147/) (2026)
- [SIA: Self Improving AI with Harness & Weight Updates](https://mlsanity.com/paper/2605.27276/) (2026)
- [Test-Time Adaptation for LLM Agents via Environment Interaction](https://mlsanity.com/paper/2511.04847/) (2025)
- [SHE: Trajectory-driven Safety Harness Evolution for LLM Agents](https://mlsanity.com/paper/2608.09885/) (2026)
- [TokenSqueeze: Performance-Preserving Compression for Reasoning LLMs](https://mlsanity.com/paper/2511.13223/) (2025)
- [Beyond Templates: Dynamic Adaptation of Reasoning Demonstrations via Feasibility-Aware Exploration](https://mlsanity.com/paper/2505.20700/) (2025)
- [Self-Evolving Coding Agents](https://mlsanity.com/paper/2608.03392/) (2026)
- [LiveThinking: Enabling Real-Time Efficient Reasoning for AI-Powered Livestreaming via Reinforcement Learning](https://mlsanity.com/paper/2510.07685/) (2025)

## Built on by (most cited first, 1 of 1)

- [A Responsive Present, a Shared Past, a Social Other: Teens' Overreliance on Companion AI Chatbots](https://mlsanity.com/paper/2609.14843/) (2026)

## Abstract (arXiv)

AI-powered digital avatar streamers must answer product questions, engage viewers, and execute marketing strategies in real time, demanding low latency, frequent strategy updates, and accurate yet effective responses. Evolvable Harnesses, whose Skills, Hooks, prompts, and tools can be updated independently of model weights, enable rapid iteration but expose a trade-off: large models adapt zero-shot yet are too slow, whereas compact models meet latency targets but overfit to fixed Harness configurations. We propose Harness-Aware Training (HAT), which trains compact models to adapt to changing Harnesses. Its key component, Harness-State Augmentation (HSA), applies task-preserving transformations to Skill identifiers and content, tool schemas, prompt structures, and Hook functions. Training proceeds in three stages: HSA-SFT learns reasoning and tool use from strong-model trajectories across diverse environments; General On-Policy Distillation restores generalization lost during SFT; and HSA-RL improves robustness to changing Harnesses through reinforcement learning in augmented environments. Across four evaluation sets, HAT achieves 94.8 on Live-Stream QA (base: 80.3; strongest general LLM: 93.0) and 94.6 on Harness-Variant QA (base: 75.4). Unlike Fixed-Harness SFT, which lowers IFEval by 7.7 points from the base model, HAT avoids this regression and reaches 83.5. On one NVIDIA H20 GPU, the optimized system delivers P50 and P95 latencies of 3.4 s and 8.1 s. Deployed in Taobao Live's digital-avatar service, it also yields positive online A/B test results for item-page views.