Yiqun Zhang

HarmTrace reveals that fine-grained target identification can drastically improve harmful meme detection accuracy, closing a critical gap in multimodal AI understanding.

Verified data synthesis can dramatically elevate the skill-use performance of language models, producing thousands of executable trajectories that enhance agent capabilities.

CAGI achieves superior imputation accuracy by leveraging latent subgroup structures, outperforming traditional methods that ignore population heterogeneity.

Achieving trillion-parameter performance with just 35 billion parameters by scaling agent horizons reveals a new frontier in model efficiency.

AGREE achieves superior clustering performance by effectively balancing attribute interaction and graph topology, overcoming the limitations of traditional methods.

LLM-based agents can now autonomously enhance their own harnesses, leading to performance boosts of up to 18% without human intervention.

Current open-world semi-supervised learning methods fall short in practical applications because they fail to extract latent semantic information, but SECOS overcomes this by directly predicting textual labels from a candidate set, achieving state-of-the-art results.

Structured composition unlocks significantly better agent performance compared to flat skill invocation, even with the same skill set.