Crimson AI NewsA CrimsonLingua Network service
EN ع
← Back to news
Research paper Hugging Face

Hugging Face Paper: Harness-Aware Training Boosts Compact Avatar Agents

AI By Crimson AI Hugging Face Papers 29 August 2026 · 00:00 5 views
Share: X Telegram

A new technical report from Hugging Face introduces Harness-Aware Training (HAT), a method that lets compact AI agents adapt to evolving digital-avatar harness configurations with low latency and high accuracy, achieving strong results in live-stream QA and deployment on Taobao Live.

Hugging Face Paper: Harness-Aware Training Boosts Compact Avatar Agents

Key points

Hugging Face has published a technical report detailing a new training approach for AI-powered digital avatar agents, particularly those used in live-streaming e-commerce. The paper, titled "Training Agents to Evolve with Their Harness," addresses the challenge of balancing low latency with the ability to adapt to frequently updated system configurations, known as harnesses.

The proposed method, Harness-Aware Training (HAT), is designed for compact models that must meet strict latency requirements. Unlike larger models that can adapt to new harnesses without retraining but are too slow, compact models often overfit to a fixed harness configuration. HAT introduces Harness-State Augmentation (HSA), which applies transformations to skill identifiers, tool schemas, prompt structures, and hook functions during training, simulating a variety of harness states.

The training process consists of three stages: HSA-SFT (supervised fine-tuning with augmented harness states), General On-Policy Distillation to restore generalization, and HSA-RL (reinforcement learning) to improve robustness. The authors report that HAT achieves a score of 94.8 on Live-Stream QA, outperforming both the base model (80.3) and the strongest general LLM (93.0). On Harness-Variant QA, it reaches 94.6, compared to the base model's 75.4.

Importantly, HAT avoids the regression seen with Fixed-Harness SFT, which lowers IFEval by 7.7 points from the base model; HAT instead reaches 83.5. The optimized system, deployed on a single NVIDIA H20 GPU, delivers P50 and P95 latencies of 3.4 seconds and 8.1 seconds respectively. The system has been deployed in Taobao Live's digital-avatar service, yielding positive online A/B test results for GMV and item-page views.

MetricHATBase ModelStrongest General LLM
Live-Stream QA94.880.393.0
Harness-Variant QA94.675.4-
IFEval83.5--
P50 Latency (s)3.4--
P95 Latency (s)8.1--
Source
Hugging Face · Hugging Face Papers
Related news
Research paper
Hugging Face 29 Aug 2026

Hugging Face Audit: 110 of 124 AI Evaluations Fail to Support Their Claims

A new commit-bound census of 124 Inspect Evals units reveals that 110 stop before deterministic inference due to missing historica...

3
Research paper
Hugging Face 29 Aug 2026

Aphanta: New Framework Diagnoses When Image Editing Boosts Multimodal Reasoning

Hugging Face researchers introduce Aphanta, a diagnostic framework that evaluates when image-editing intermediates improve multimo...

5
Research paper
Hugging Face 29 Aug 2026

EditaLive! Enables Real-Time Character Video Editing for Live Streaming

Hugging Face researchers introduce EditaLive, a framework for real-time human-centric video editing in live streams, achieving sta...

3