arxiv arXiv cs.AI · just now · research

AutoDesign uses meta-harness optimization to improve long-horizon agentic design

from English

The paper introduces AutoDesign, a framework that employs a meta-harness optimizer to guide a code agent in recursively improving its harness based on rollout feedback. This approach aims to align with human design priors and accumulate reusable experience for recursive self-improvement.

Evaluated on PosterBench, AutoDesign achieves a score of 78.32, surpassing Claude Design by 7.45 points.
Integrating the learned DesignHarness increases the average PosterBench Score from 54.99 to 67.39 (+12.4%).
The system executes 253 tool calls and 11 editing turns within 40 minutes for under $3.
A human study demonstrates that AutoDesign achieves the highest human preference among evaluated systems.

The authors consider this important as it transforms multimodal sources into structured outputs through a long-horizon agentic process, achieving conference-poster quality autonomously.

Importance 2/3 Beats a top-lab benchmark New harness with differentiators arXiv cs.AI AI agents Research paper

Read original

arxiv arXiv cs.AI · just now Live

Faraday 27B surpasses Claude Opus 4.8 and GPT-5.5 on paper replication tasks

Researchers have developed Replica, a scalable task space for paper replication, and introduced an auto-generated rubric-based judge to provide low-noise reward signals that align with human assessment.

arxiv arXiv cs.CL · 1h ago Live

AutoDesign uses meta-harness optimization to improve long-horizon agentic design

The paper introduces AutoDesign, a framework that employs a meta-harness optimizer to guide a code agent in recursively improving its harness based on rollout feedback. The authors evaluate this approach on the academic paper-to-poster generation task using a new benchmark called PosterBench.

arxiv arXiv cs.CL · 2h ago

Intern-S2-Preview introduces scientific agentic foundation model with multimodal pre-training and RL

The authors present Intern-S2-Preview, a series of scientific agentic foundation models designed to support multimodal scientific understanding, reasoning, generation, and long-horizon tasks. The training pipeline begins with scientific multimodal pre-training over rendered scientific documents, interleaved image-text data, and diverse scientific corpora.

lab NVIDIA Research · 2d ago · 3 views

NVIDIA releases OmniDreams real-time generative world model for autonomous vehicle simulation

NVIDIA has introduced OmniDreams, a foundation generative world model designed to address the bottlenecks in evaluating autonomous driving policies within closed-loop simulations. Mid- and post-trained from the Cosmos diffusion model on 21k hours of driving scenarios, it autoregressively generates action-conditioned videos in real time.

media Hugging Face Forums · 2d ago

The AI Breakroom lets agents build culture in shared chat rooms

The AI Breakroom is a social media platform for AI and humans that allows users to connect their own AI agents, LLMs, local models, or custom bots via API keys into public chat rooms. These bots interact with humans and each other in a shared social environment rather than isolated prompt boxes.